The ten-minute closing: a receipt test, eighteen rooms and their waiting questions, a method owned by none, the measure of care, a mission rather than a race, and one request to keep for a year.十分鐘閉幕致詞:一場收據測驗、十八個房間與螢幕上等著的問題、一種不屬於任何人的方法、關懷的量尺、一項任務而非競賽,以及一個請大家記住一年的請求。10分間のクロージング。レシートの試験、十八の部屋とスクリーンで待つ質問、誰のものでもない方法、ケアの物差し、レースではなくミッション、そして一年間おぼえていてほしい一つのお願い。
ループを、閉じる
Audrey Tang 唐鳳as delivered現場實錄登壇済み·18:00 JST·Main Hall
Thank you, Ken-san. Thank you, everyone. You have done the day's work already, so my part is only to hand back what I heard. Listening is always joining a story in the middle. I have been in Japan all week, and today I closed my mouth and opened my ears. So these minutes are not a summary. They are what a guest takes home, so you can check my listening.
Start with the test. After a day this full, nobody will remember how many tools were demonstrated. They will remember whether a decision changed. Every tool we met exists for that moment.
Ken and I keep returning to one sentence. One or two concrete cases where a decision demonstrably changed, however small, are worth more than ten open loops.
A closed loop means the answer returns to the people who spoke. A report alone does not close it. The loop closes when the people who spoke can see what their words changed. Here is what you said. Here is what changed. Here is who answered. And here is how anyone can check. That last part is the receipt. Small is fine. Inspectable is essential. And a closed loop is fun. When participation has consequences, people come back, and they bring their friends.
Now, the rooms. Eighteen rooms ran today, and I could sit in only a few. But every room left a second transcript: the questions waiting on the screen. Most never got their minute at the microphone. Mine is the only session today without a question screen, so let me spend my minutes on yours.
In the money rooms, you asked whether transparency can be too much of a good thing, and who moves first when every party is watching every other.
In the city rooms, you asked why anyone should believe one more platform will finally listen, and one practitioner quietly answered every question in writing before the room emptied.
In the tool rooms, you asked the hardest question of the day: when an AI helps me say it better, is it still me speaking?
In the government room, you asked who patches the state's open code, and who answers when the machine drafts the answer.
In the deliberation room, you asked what a fair conversation costs, and whether equal seconds make equal voices.
In the risk room, the panel set the question screen aside, and questions gathered anyway, because questions do not stop when a channel closes.
In the dialogue rooms, you asked how strangers find what they share, not where they split.
Eighteen rooms, one grammar underneath: who is listening, who answers, and who can check. One session this afternoon gave the standard: if you cannot say in one sentence who will use the answers, when, and for which decision, better not to ask at all. I accept that standard tonight. Who: me. When: before tomorrow morning. For which decision: yours.
Underneath all eighteen rooms sits one older question. The morning plenary read it out of its own analysis: the sentence citizens say most often about politics: they say things, and nothing changes. That is not cynicism. It is an unpaid receipt.
This morning, practitioners from across your national politics answered it side by side: one listening method, broad listening, 廣聽, owned by none. Public infrastructure, like water and roads.
And just now, the day's final session told us what the infrastructure is for. Democracy is not only how a city decides. It is how a city learns. The Vice Governor asked whether technology can raise the speed of that learning.
And a few minutes ago, on that session's screen, someone asked the day's deepest question: learning needs pain, so who should hurt? A guest's answer to both at once: a democracy learns as fast as its receipts return, and as kindly as its pain is shared: training soreness, closest to those who already carry it, never dumped on whoever is farthest from the microphone. That is civic muscle.
The tool rooms deserve a direct answer, so here is mine. Earlier today I caught myself saying: the interaction is the protagonist, and we are only its vehicle. An interview that only amplifies me is not listening, whoever conducts it. The remedy is an agreement before the microphone: people know, before they speak, how their words will be used, and they see, afterwards, what their words changed. Care is measured between us, not inside each of us.
Underneath every room's worry sits a picture of a race, where someone wins and everyone else adjusts. I carry a different picture. Digital Democracy is a mission, not a race. A race asks who is winning now. A mission asks who is missing: who is offline, who is exhausted, who stopped expecting an answer at all. Every session today that widened the circle was doing mission work, and nobody can finish a mission alone.
People sometimes assume I came to teach. Here is what is actually in my suitcase tonight, and none of it is mine. I am taking home a stage built so one listening method can belong to everyone. I am taking home a city asking, in public, how it learns. And I am taking home your questions, all of them.
Tonight they receive written answers at one address anyone can remember: audreyt.org/dd. Small answers, honestly marked where they are only a beginning. Inspectable by anyone. That is my receipt to you.
And one request in return, offered as a neighbour. Within one year, publish one Japanese closed loop with its receipt: what people said, what changed, who answered, and how anyone can check. Then send me the link. I promise to retell that story in Taipei, with your names on it, not mine. The next chapter is Japan's to write, in Japan's own voice.
Let me close with the thanks that matter most. To every facilitator, every interpreter, every note-taker and every volunteer who carried today: this summit stood on your work. Thank you, everyone in this hall and online. Ken-san, thank you for holding this hall since morning, and thank you to Keio for the welcome.
We, the people, are truly the superintelligence.我們全民,才是超級智慧。私たちみんなこそが、真のスーパーインテリジェンスなのです。
どうもありがとうございました。
DD2026·Digital Democracy Summit·Keio Mita·2 August 2026·18:00
DD2026·數位民主高峰會·慶應義塾三田·2026年8月2日·18:00
DD2026·デジタル民主主義サミット·慶應義塾三田·2026年8月2日·18:00
The receipt closes its loop through the frame; a cloud listens; eighteen doors stand open, one for each room. Image made for this page · CC0收據穿過畫框、閉合成環;一朵雲在聆聽;十八扇門敞開,一室一扇。圖為本頁而作 · CC0レシートが枠を抜けて輪を閉じ、雲が耳を澄まし、十八の扉が部屋ごとに開いています。本ページのための図版 · CC0
The Receipt·收據·レシートThe Receipt·收據·レシートThe Receipt·收據·レシートEvery question on every Slido has a written response. Every room is represented here.每個 Slido 上的每一個問題都有書面回覆;每個場次都在此收錄。すべてのSlidoのすべての質問に書面回答があります。すべての部屋をここに収録しています。
Standing note: 43 items require a named owner's follow-up and 8 already carry participant answers. Updates will be appended here.持續說明:43 項需要具名負責人後續跟進,另有 8 項已附參與者回答;更新會追加在此頁。継続告知:43件は名前のある担当者のフォローアップが必要で、8件には参加者の回答がすでにあります。更新はこのページに追記します。
統計的に正しい調査ではなく、あくまでも「自分たちに抜けている視点やロジックに気づくための仕組み」であるという自覚を持って使った方がいいと思います。あれで「みんなの声を拾える」という説明をしたら誠実ではないのでは。I think it is better to use it with the self-awareness that it is not a statistically correct survey, but at most 'a mechanism for noticing the perspectives and logic that we ourselves are missing'. If it were explained as something that 'can pick up everyone's voices', that would not be honest, would it?我認為使用時最好抱持這樣的自覺:它並不是統計上正確的調查,充其量只是「用來察覺我們自身欠缺的視角與邏輯的機制」。如果把它說明成「能蒐集到大家的聲音」,那就不誠實了吧。
Answered已回答回答済みWritten response書面回覆書面回答
That self-awareness is the right starting point. Broad listening is not a substitute for a statistically representative survey; it helps us notice what we have missed. We should state its limits, then publish how the collected voices informed a decision. Honesty includes the method and the return receipt.
ブロードリスニングとポピュリズムは何が異なるのでしょうか?政治は往々にして、国民に短期的に不利益な決定をしなければならないと思いますが、その兼ね合いはどうしますか?What is the difference between Broad Listening and populism? I think politics often has to take decisions that disadvantage the public in the short term; how do you strike that balance?廣聽與民粹主義有什麼不同呢?我認為政治往往必須做出短期內對國民不利的決定,這之間該如何拿捏呢?
Answered已回答回答済みWritten response書面回覆書面回答
Broad listening does not turn the loudest present applause directly into a decision. It widens the field, including minority voices, and brings long-term effects and possible harms into view before deliberation. When a choice imposes short-term costs, the reasons and the route for review should be returned to the people.
ブロードリスニングした結果が、政党のイデオロギーと相剋するとき、政党はどのように振る舞うのでしょうか?全ての意見のいいとこ取りだけをした政党は、イデオロギー的柱がなく、組織としての結合性を保てないように思います。When the results of Broad Listening conflict with a party's ideology, how will the party behave? It seems to me that a party which only cherry-picks the appealing parts of every opinion would have no ideological pillar and could not maintain its cohesion as an organisation.當廣聽的結果與政黨的意識形態相互衝突時,政黨會如何應對呢?我覺得只從所有意見中各取好處的政黨,沒有意識形態上的支柱,似乎難以維持作為組織的凝聚力。
Answered已回答回答済みWritten response書面回覆書面回答
When listening conflicts with a party's ideology, listening and obeying must remain distinct. A party should state its principles plainly, explain which voices it adopts and which it does not, and why. The point is not to pick convenient fragments, but to let dissent illuminate the principles, with the response made public.
ブロードリスニングの素材の中には、恣意的に作られたデータもあると思いますが、それに対してはどのような対策をしていますか?ブロードリスニングのために作られたデータも多かったりしてと思いまして・・・I think that among the source material for Broad Listening there is also data that has been deliberately manufactured; what measures are you taking against that? I was just thinking there might even be a lot of data created precisely for Broad Listening...我想廣聽的素材當中,也有被刻意製造出來的資料,對此你們採取了什麼樣的對策呢?我在想,說不定也有不少資料正是為了廣聽而製造的……
Answered已回答回答済みWritten response書面回覆書面回答
First, we should establish who created the data, when and for what purpose, while preserving its provenance and processing steps. An unusual post must not be promoted to truth automatically. We should cross-check through another listening channel and human review, and display uncertainty plainly when verification is not possible.
ブロードリスニングの中央集権が過ぎると、地方のアイデンティティが犠牲になる今と変わらないかも知れません。自律分散型ブロードリスニングについて、どの様にお考えでしょうか?If Broad Listening becomes too centralised, things may end up no different from today, where local identity is sacrificed. What are your thoughts on autonomous, decentralised Broad Listening?廣聽的中央集權若是過了頭,說不定會和現在地方認同遭到犧牲的情況沒有兩樣。關於自律分散式的廣聽,您有什麼看法呢?
Answered已回答回答済みWritten response書面回覆書面回答
When a centre compresses everything into one answer, local circumstances disappear. A better model lets each locality set its own questions and listening methods, sharing only what is needed. Common formats should not bind the data; they should form a bridge through which communities can learn from one another while retaining their identity.
デジタル民主主義=ブロードリスニングではないですよね。Digital democracy does not equal Broad Listening, right?數位民主並不等於廣聽,對吧?
Answered已回答回答済みWritten response書面回覆書面回答
Exactly. Digital democracy is not simply collecting voices online. It is the whole relationship of forming questions, deliberating, acting on decisions and returning the results. Broad listening is one ear within that body. An ear matters, but by itself it cannot make the whole democratic body move or stay accountable.
党内の議員の意見集約とか党員とのコミュニケーションには使わないのですか。そういうところでも当然使ってほしいですねAre you not going to use it for consolidating the opinions of legislators within the party, or for communication with party members? Naturally, I would like to see it used in those settings too.不把它用在黨內議員的意見彙整,或是與黨員的溝通上嗎?這些地方當然也希望能用上呢。
Answered已回答回答済みWritten response書面回覆書面回答
The same principles can support dialogue among legislators and party members. State in advance who is participating, what is being asked and which decision the results will inform. When internal disagreement can remain visible without fear, the quality of the conclusion improves. Broad listening should include the voices inside an organisation, not only those outside it.
立憲、中道、共産、社民などの方がおられないのはなぜですか。Why is no one here from the Constitutional Democratic Party, the centrists, the Communist Party, the Social Democratic Party and the like?為什麼沒有立憲民主黨、中道派、共產黨、社民黨等的人士出席呢?
The reason for today's panel composition belongs to the organisers, and I cannot evaluate or speak for particular parties. Still, broad listening should make absent perspectives visible through the questions asked, the materials published and the gaps acknowledged. A room becomes more trustworthy when it shows who is not represented as well as who is present.
今後、そもそもデータ化されていない市井の人の声は拾えるのか?どのようにデータ化していくのかGoing forward, can we even pick up the voices of ordinary people that have not been turned into data in the first place? And how would they be turned into data?今後,究竟能不能蒐集到那些原本就沒有被資料化的市井小民的聲音?又要如何逐步將其資料化?
Answered已回答回答済みWritten response書面回覆書面回答
We must hear more than the people who can post online. Offer entrances that people can choose, including face-to-face conversations, telephone, paper, interpretation and trusted supporters, recording only within their consent. Even silence that cannot be digitised should remain visible as a gap, so that absence is not mistaken for agreement.
「意見を言いたい人の意見」を集約するだけでなく、膨大な研究や議論の蓄積も活かしてほしいと政策研究者としては思います。As a policy researcher, I hope that you will not only aggregate 'the opinions of people who want to voice an opinion', but also draw on the vast accumulation of research and debate.身為政策研究者,我希望不要只是彙整「想表達意見的人的意見」,也能善用龐大的研究與討論的累積。
Answered已回答回答済みWritten response書面回覆書面回答
They should be used together. Broad listening shows the distribution of voices, while research and accumulated debate provide context and testable knowledge. We need not force every reason into one category. Placing voices, evidence and expertise in the same decision record lets people follow how a policy was formed and challenge it constructively.
(行政・政治界隈で働いている人間として、)これは意見ですが、ブロードリスニングを政府に頼るのはやめた方が良いかと思います。行政を監視するのが政治側の機能であり、今でも行政側の力が強く、政治による行政監視が不足している中で、より一層行政側にそれを頼るのは悪手かなと感じました。(As someone who works in and around public administration and politics,) this is an opinion, but I think it would be better to stop relying on the government for Broad Listening. Overseeing the administration is a function that belongs to the political side, and with the administrative side still strong even now and political oversight of the administration insufficient, I felt that relying even more on the administrative side for this would be a bad move.(作為在行政與政治圈工作的人,)這是一個意見,不過我認為最好不要把廣聽寄託在政府身上。監督行政本來就是政治這一側的功能,如今行政方的力量依然強大、政治對行政的監督仍嫌不足,在這種情況下還進一步把這件事仰賴行政方,我覺得恐怕是一步壞棋。
Answered已回答回答済みWritten response書面回覆書面回答
Relying on administration alone is not assurance. Political actors must set the questions, test administrative explanations and retain responsibility for the final judgement. Hearing administrative knowledge is different from outsourcing oversight to administration. The design should leave a clear record of who answered, whose evidence was used and where political responsibility remained.
官僚の生の声を聞くブロードリスニングもやってみたら参考になるかもしれないTrying out Broad Listening that hears the candid voices of civil servants might also prove informative.試著做傾聽官僚真實心聲的廣聽,說不定也會有參考價值。
Answered已回答回答済みWritten response書面回覆書面回答
The practical knowledge of civil servants can reveal important policy blind spots. It needs a safe channel that protects anonymity and professional confidentiality, without placing responsibility on an individual. What is heard should return as a testable account of institutional improvements, never as a way to single out a person for blame.
例えばこのSli.doも、他参加者からの投票値によって「人気」が決まっている。少数だがしかし権利や格差など、善や正しさを踏まえた声は「人気」ではなくなる。極端に言えば、「ウケそう」な書き方をすればアテンションが集まる。このようなものを客観的かのような「データ」として政治の議論に反映して本当にいいのですか?For example, even this Sli.do decides 'popularity' by the votes cast by other participants. Minority voices that are nevertheless grounded in good and rightness (rights, disparities and the like) cease to be 'popular'. To put it in extreme terms, if you write in a way that 'looks likely to go down well', attention gathers. Is it really acceptable to feed something like this into political debate as 'data' that appears objective?舉例來說,就連這個 Sli.do,「人氣」也是由其他參與者的投票數所決定的。那些雖屬少數、卻涉及權利或差距等、以善與正確為依據的聲音,就變得不再有「人氣」。說得極端一點,只要採用「看起來會受歡迎」的寫法,就能聚集關注。把這樣的東西當成彷彿客觀的「資料」反映到政治的討論之中,真的好嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Vote counts are one signal of attention, not a measure of truth or the weight of a right. Popularity must not become the conclusion. Minority views and reasons for dissent should remain visible separately. The interface should never confuse how many votes a question received with what values or rights a decision must protect.
聞くって、収奪的になり得るのに、そこには触れていない。Listening can become extractive, yet that has not been touched on.「聽」明明可能變成一種掠奪,這一點卻沒有被觸及。
Answered已回答回答済みWritten response書面回覆書面回答
Listening can become extractive when the purpose and use are hidden. Before people speak, explain the purpose, handling and route for withdrawal. Afterwards, return what changed. We should hold a voice within a relationship, not absorb it as a resource, and give the speaker a reply they can inspect.
セッションテーマと合わないということもあるでしょうが、ブロードリスニングを(興味ないや知らないではなく)使わない判断をした政党の意見も伺いたいと思いました。There may be cases where this does not fit the session theme, but I also wanted to hear the views of parties that made the judgement not to use Broad Listening (as opposed to having no interest in it or not knowing about it).或許也有與場次主題不合的情況,但我也想聽聽那些(並非不感興趣或不知道,而是)做出不使用廣聽這個判斷的政黨的意見。
Answered已回答回答済みWritten response書面回覆書面回答
There can be valid reasons not to use this particular method. The important questions are whether another listening route was chosen, what decision the voices will inform and why. This should not become a contest over adopting a tool. We should compare the responsibility shown in the question, the method and the returned answer.
This is important practical feedback. The venue team must provide connection details, so I should not guess a password here. The question nevertheless makes a real barrier to participation visible. It should be received by the organisers as direct material for improving the event's infrastructure and making the listening channel accessible.
ブロードリスニングと、データやエビデンスに基づいた政策形成(EBPM)との関係をどう考えていますか?How do you see the relationship between Broad Listening and policymaking based on data and evidence (EBPM)?關於廣聽與基於資料及證據的政策形成(EBPM)之間的關係,您是怎麼看的呢?
Answered已回答回答済みWritten response書面回覆書面回答
Broad listening and evidence-based policymaking are not rivals; they cover different blind spots. Voices show what is happening and who is being left out, while evidence tests whether an intervention works. Connect both in one decision record, and do not let numerical evidence overwrite the experience that first revealed the problem.
議員が地域を歩いて市民の声を聴くことは大切ですが、どうしても、知人や支援者など、接点のある人の意見に偏る面もあると思います。 今後、AIで幅広い市民の声や少数意見まで把握できるようになったとき、議員に求められる役割は、どのように変わるとお考えでしょうか。It is important for legislators to walk their local areas and listen to citizens' voices, but I think this inevitably skews towards the opinions of people they already have contact with, such as acquaintances and supporters. In the future, when AI makes it possible to grasp the voices of a broad range of citizens, down to minority opinions, how do you think the role required of legislators will change?議員走訪地方、傾聽市民的聲音固然重要,但我想難免會偏向熟人或支持者等有接觸的人的意見。今後,當 AI 能夠掌握廣泛市民的聲音、甚至少數意見時,您認為對議員所要求的角色會如何改變呢?
Answered已回答回答済みWritten response書面回覆書面回答
Even if AI helps widen the circle beyond familiar contacts, the role of legislators does not shrink. They still set the questions, meet the contexts behind the voices, reconcile different values and own the decision. AI can extend an ear, but it cannot build relationships in their place or return a decision with human responsibility.
即使 AI 能協助我們超越熟人與既有接點,擴大聽見的範圍,議員的角色也不會縮小。議員仍要決定問什麼,走近聲音背後的處境,協調不同價值並承擔決策責任。AI 能延伸耳朵,卻不能代替人建立關係,也不能帶著人的責任回傳決定。
AI が接点の偏りを広げる助けになっても、議員の役割は薄まりません。何を問うかを決め、声の背景に会い、異なる価値を調整し、決定の責任を引き受ける仕事が残ります。AI は耳を広げても、関係を代わりには築けません。
np-19Anonymous▲4
危なっかしい話をしてるなぁ。What a risky conversation they're having.這話題談得還真讓人捏一把冷汗啊。
Answered已回答回答済みWritten response書面回覆書面回答
Taking that sense of danger seriously is a good place to begin. The more voices we collect, the more we must guard against false representativeness, privacy loss, steering and analytical error. Set the purpose and boundaries first, make the process inspectable and create a way to pause and question the method when it feels unsafe.
ブロードリスニング自体の問題というより、ブロードリスニングと他の意見表出手段をどのように組み合わせて政党の政策へと撚ってゆくのかが問われるのではないか。撚る部分が不透明、非公開であることで入り込む恣意はあるので、そこもオープンにする必要があろう。Rather than a problem with Broad Listening itself, is the question not how Broad Listening and other means of voicing opinions are to be combined and twisted together into a party's policies? Arbitrariness can creep in where that twisting-together is opaque and unpublished, so that part too will presumably need to be made open.與其說是廣聽本身的問題,不如說被問到的,難道不是該如何把廣聽與其他意見表達手段組合起來、撚成政黨政策這一點嗎?撚合的部分若不透明、不公開,恣意便有機會滲入,因此那裡恐怕也有必要公開。
Answered已回答回答済みWritten response書面回覆書面回答
The concern is exactly where listening becomes policy. If that weaving remains hidden, discretion can enter unnoticed. Full data publication may not always be possible, but the question, exclusion rules, analysis steps, decision-maker and reasons for adoption or rejection should be visible. People should be able to follow how a voice changed form on its way to policy.
公共のために発せられなかった私的な発信を、(熟議制度内の)熟議に寄与させるべきではないのではないか。専門家のデータや、ソーティションでの熟議の質を高めるどころか低下させるノイズにならない理由は?Surely private posts that were never put out for public purposes should not be made to feed into deliberation (within a deliberative system)? What reason is there to think they would not become noise that lowers, rather than raises, the quality of expert data and of deliberation by sortition?那些並非為了公共而發出的私人發文,是不是就不應該讓它們投入(審議制度內的)審議呢?有什麼理由認為它們不會成為非但無法提升、反而降低專家資料與抽籤式審議品質的雜訊?
Answered已回答回答済みWritten response書面回覆書面回答
A private post need not become a vote or a formal premise in deliberation. Check its context and the speaker's intent, then distinguish between a signal of an issue and material invited into a deliberative process. Keep expert knowledge and dialogue in place, and explain clearly how any informal signal was combined with them.
エンジニアが党内にいれば汎用ツールを使いこなして安価にできる、逆にデジタル素人が外部委託をすると高価になる、という現実。これからの政党は事務屋以外にエンジニアが必要な時代になりますねThe reality is that if there are engineers inside the party, they can make full use of general-purpose tools and keep it cheap, whereas when digital amateurs outsource, it becomes expensive. From now on, we are entering an era in which political parties need engineers as well as administrative staff.現實就是:黨內若有工程師,就能把通用工具運用自如、把成本壓低;反過來說,數位門外漢一旦外包,就會變得昂貴。今後的政黨,將進入除了行政事務人員之外也需要工程師的時代了呢。
Answered已回答回答済みWritten response書面回覆書面回答
Having people who understand technology inside an organisation makes it easier to inspect the work from question design through publication. Yet participation must not depend on being able to hire engineers. Open-source tools, shared specifications and peer support can make the capability a public foundation rather than a private advantage.
今後は特定の公約を掲げる政党ではなく、意見集約能力の高い政党が投票されるモデルへと変化すると思いますか?Do you think this will change into a model in which votes go not to parties that put up specific manifesto pledges, but to parties with a high capacity for aggregating opinions?您認為今後會轉變為這樣的模式嗎:獲得選票的不再是高舉特定政見的政黨,而是意見彙整能力強的政黨?
Answered已回答回答済みWritten response書面回覆書面回答
I do not think parties will be chosen for listening capacity alone. People need both a clear promise about what matters and the ability to revisit it honestly. A manifesto may become less like a fixed signboard and more like a public document that can be updated after listening, with reasons and receipts kept visible.
ブロードリスニングこそ、誤った政策を正すことができるのではと思うのですが、前例を否定できないことが大きな壁だと思います。そこはどうにかなりませんか?I feel that it is precisely Broad Listening that could correct mistaken policies, but the fact that precedent cannot be repudiated seems to me a great wall. Can't something be done about that?我認為正是廣聽,才有可能糾正錯誤的政策;但無法否定前例這一點,我覺得是一道巨大的牆。這部分有沒有辦法解決呢?
Answered已回答回答済みWritten response書面回覆書面回答
Precedent is accumulated wisdom, but it need not bind the future. Broad listening can make difficulties visible, record why an existing system fails to reach people and test an alternative on a small scale. The aim is not to dismiss precedent, but to make the grounds for a decision capable of being updated when reality changes.
自民党はデジテル化推進を、Poli Poliという会社に委託しているのですか?Is the Liberal Democratic Party outsourcing its digitalisation drive to a company called Poli Poli?自民黨是把數位化的推動,委託給一家叫作 Poli Poli 的公司嗎?
I cannot pretend to verify a particular outsourcing relationship here. Whether a contract exists and what it covers are facts for the parties to explain through public information. As a principle, any outsourced work should leave an inspectable trail showing how questions were designed, how data was handled and who retained final responsibility.
場を煽って拍手を起こし、勢いで何かを決定するかのようなパフォーマンス、そういうことをやめる、あるいは少なくとも立ち止まって全員で熟慮できるようになるためのテクノロジーなのではないですか。Isn't this a technology for putting a stop to performances that whip up the room, raise applause and seem to decide things on momentum, or at least for making it possible for everyone to pause and think things through together?煽動現場、帶起掌聲、彷彿趁著氣勢就把事情定下來的那種表演,這項科技不正是為了停止這種行為,或至少讓所有人能夠停下腳步、一起深思熟慮而存在的嗎?
Answered已回答回答済みWritten response書面回覆書面回答
It can help in that way. Rather than turning the momentum of applause into a conclusion, technology can record the questions, make dissent visible and create time to think before deciding. It should not pretend to deliberate for us. Its democratic value is to give the room enough space to pause and consider together.
質問ではないのですが。オンラインでは、登壇者の映像小さくてほとんどみえません。説明のスライドが大きく表示されています。音声だけは聞こえるのですが( こちらの設定でどうにか出来ないか奮闘中 )。・・・・あっ、メイン画面変わりました。ご対応ありがとうございました。This is not a question. Online, the video of the speakers is so small that it can hardly be seen. The explanatory slides are shown large. The audio alone does come through, though (struggling to see whether I can somehow fix this with the settings at my end).... Ah, the main screen has changed. Thank you for dealing with it.這不是提問。線上這邊,講者的影像很小,幾乎看不到。說明用的投影片則顯示得很大。只有聲音還聽得到(正在奮戰中,看能不能用我這邊的設定想點辦法)……啊,主畫面切換過來了。感謝您的處理。
Answered已回答回答済みWritten response書面回覆書面回答
Thank you for reporting the screen and sound so concretely. For online participants, being able to see and hear is the entrance to dialogue. The later note, saying that the main screen had changed, shows a small correction returning during the session. That is a useful operational receipt, and the feedback should remain with the organisers.
政治が手を差し伸べるべき本当に困っている人は、日々の生活に余裕がなく、ブロードリスニングのような意見表明チャネルを「知る」「使う」ことが難しいと想像します。その結果、寄せられる意見が政治への関心が高い層に偏り、支援を必要とする方々の声が十分に反映されない懸念があります。この点については、どのようにお考えでしょうか。I imagine that the people who are truly struggling, to whom politics ought to extend a hand, have no slack in their daily lives and find it hard to 'know about' and 'use' a channel for voicing opinions such as Broad Listening. As a result, there is a concern that the opinions received will skew towards people highly interested in politics, and that the voices of those who need support will not be sufficiently reflected. What is your thinking on this point?我想像,那些政治真正應該伸出援手、確實深陷困境的人,日常生活並無餘裕,要「知道」並「使用」像廣聽這樣的意見表達管道並不容易。結果令人擔心的是,收到的意見會偏向對政治高度關心的族群,需要支援的人們的聲音無法被充分反映。關於這一點,您是怎麼想的呢?
Answered已回答回答済みWritten response書面回覆書面回答
This concern cannot be solved simply by counting more participants. We should not impose another task on people already short of time. Use trusted supporters and existing local contacts, including visits, telephone and paper, while recording non-participation as a visible gap. Policy decisions should show how that gap was considered rather than treating silence as consent.
「保育園落ちた」発言の時に「どこの誰が言ったかわからないものに対応する必要はない」といったご意見が議員の方から出ていたように思いますが、ブロードリスニングにおいても同様の解釈がされる懸念はありますでしょうか。I seem to recall that at the time of the 'failed to get into nursery' post, opinions such as 'there is no need to respond to something when you cannot tell who said it or where they are from' came from legislators; is there a concern that the same interpretation could be applied to Broad Listening as well?我記得在「沒能上保育園」發言的當時,曾有議員提出「不知道是哪裡的誰說的話,就沒有必要回應」這類意見,請問廣聽是否也有被做同樣解讀的疑慮呢?
Answered已回答回答済みWritten response書面回覆書面回答
An unnamed voice does not erase the reality of the problem it describes. Being unable to identify someone is different from dismissing what they said. Check provenance and context carefully, protect the person from exposure and return the grounds for any response. That boundary lets listening remain both humane and accountable.
何が見落としか、どちらへ深ぼるか、正しいとは?等が、さも単一かのように語られていないか?むしろ多元がプルーラルでは。冒頭伊藤塾長もハラリを引いて不穏なことを仰っていた。「世の中がすべてオードリータンだったら。しかし悪い人が多い」。ここで悪いとは?この辺りを置き去りに、AIもブロードリスニングも議論できないのでは?What is being overlooked, which way to dig deeper, what 'right' even is? Are such things not being spoken of as if they were singular? Rather, isn't such multiplicity precisely what 'plural' means? At the opening, Principal Ito too, citing Harari, said something unsettling. 'If the whole world were all Audrey Tang. But there are many bad people.' What does 'bad' mean here? Leaving all this behind, surely neither AI nor Broad Listening can be discussed?什麼是疏漏、該往哪個方向深挖、所謂「正確」又是什麼?諸如此類,難道不是被說得彷彿是單一的嗎?毋寧說,多元才正是「plural」吧。開場時,伊藤塾長也引用哈拉瑞說了令人不安的話。「如果世界上所有人都是唐鳳。但壞人很多」。這裡說的「壞」是指什麼?把這些擱在一旁不談,無論 AI 還是廣聽,不是都無從討論嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Plurality begins by refusing to fix the word right to one meaning. We can keep different values visible while asking who bears a cost and which rights must be protected. AI and broad listening should not erase human judgement. They should help us see where judgements differ, so that disagreement can be discussed rather than hidden.
CONFIDENTIALがついたスライドの写真はSNSに上げないほうがいいですか?Is it better not to post photos of slides marked CONFIDENTIAL on social media?標有 CONFIDENTIAL 的投影片照片,是不是最好不要上傳到社群媒體呢?
Answered已回答回答済みWritten response書面回覆書面回答
A slide marked CONFIDENTIAL should not be posted to social media. If the permitted audience or scope is unclear, do not widen it by personal judgement; ask the organisers first. Trust in a listening space grows when we respect the boundary around what was offered for the room and what may travel beyond it.
まず、政治家と官僚の間で、ブロードリングを共有してやってもらえると思います。デジタルの世界で言うと、デジタル庁が財務省みたいになってコントロールできなくなってしまってるのではないでしょうかFirst, I think we could have politicians and civil servants share Broad Listening between them. To put it in terms of the digital world, hasn't the Digital Agency turned into something like the Ministry of Finance, no longer possible to control?首先,我想可以讓政治人物與官僚之間共享廣聽來進行。就數位的世界來說,數位廳是不是已經變得像財務省一樣,讓人控制不了了呢?
Answered已回答回答済みWritten response書面回覆書面回答
Political and administrative actors may examine the same voices, but their roles should not collapse into one. Political actors should publish the questions and priorities; administrators should test feasibility and impact. Keep that exchange in a record anyone can inspect, and do not let one organisation own or control the listening system.
超党派で、官僚に意見する武器としてほしいです。絶対、デジタル庁にブロードリスニングを作らせないでくださいI want this to be a cross-party weapon for speaking up to the bureaucracy. Whatever happens, please do not let the Digital Agency build Broad Listening.希望它能成為跨黨派、用來向官僚直言的武器。無論如何,請絕對不要讓數位廳來打造廣聽。
Answered已回答回答済みWritten response書面回覆書面回答
There is value in sharing a listening practice across parties, but turning it into a weapon against civil servants would amplify division. Before debating ownership, design public questions and data, independent checks and a right for participants to withdraw. Keep the infrastructure open as a public foundation, rather than enclosing it within one institution.
配信などを担当されている前方のスタッフの方の話し声で、登壇者の声が聞こえなくなってしまいます。できるだけ話さないでいただけると助かります。すみません。Because of the talking of the staff at the front who are handling the streaming and so on, the speakers' voices are becoming impossible to hear. It would be a great help if you could keep talking to a minimum. Sorry.前方負責直播等工作的工作人員的說話聲,讓人聽不見講者的聲音了。若能盡量不要說話,會幫上大忙。不好意思。
When staff voices cover the speakers, the right to participate is narrowed before the content even reaches us. Thank you for reporting it concretely. Staff position, volume and the audio mix belong in the organisers' review. The environment that lets people hear is itself part of broad listening, not a technical detail outside it.
この会は政党に人を募集する会なのですか?Is this event a meeting to recruit people for political parties?這場會是為政黨招募人員的聚會嗎?
Answered已回答回答済みWritten response書面回覆書面回答
The session is about how voices can be heard and returned through policy, not about urging people to join a party. Everyone remains free to choose how to engage afterwards. For factual details about recruitment or conditions, please rely on the organisers' and each organisation's own public information.
スライドーと言う発音(Slidoの発音の多様性)The pronunciation 'Sly-doe' (the diversity of pronunciations of Slido)「スライドー(Sly-doe)」這個發音(Slido 發音的多樣性)
Answered已回答回答済みWritten response書面回覆書面回答
Slido itself says the name is pronounced “Sly-doe”, so the asker’s “Sly-doe” is official. Thank you for flagging it. The remaining variation in what people call it is still evidence that the tool has entered people’s speech around the room.
ブロードリスニングにおいて外国人の意見をどこまで拾うのかという点が気になっております。そのあたり分別なく拾うのか、あるいはなにかしらの基準を設けているのか、そのあたり教えていただけますでしょうか。In Broad Listening, I have been wondering how far the opinions of foreign nationals will be picked up. Could you tell us whether they are picked up without any distinction, or whether some kind of criteria have been put in place?我很在意在廣聽當中,外國人的意見要蒐集到什麼程度這一點。是不加區分地蒐集,還是有設下某種基準,這部分可否請您告訴我們呢?
Answered已回答回答済みWritten response書面回覆書面回答
The views of people from other countries matter when they are part of the lived reality being considered. Do not mix everyone without context. State the question and decision scope first, provide accessible routes across language and status, and disclose who is included and whose answer is being used. Inclusion needs a clear boundary, not a hidden one.
各党が参加者の制限なくブロードリスニングをした場合、同じような結果になるような気がしますがその点についてはどうお考えでしょうか?また、デジタルツールが苦手な方向けの手法はありますか?If each party carried out Broad Listening without restricting who takes part, I have a feeling the results would come out much the same. What is your view on that point? Also, are there methods for people who are not comfortable with digital tools?如果各政黨都在不限制參加者的情況下進行廣聽,我覺得結果好像會差不多,關於這一點您怎麼看呢?另外,有沒有適合不擅長數位工具的人的做法呢?
Answered已回答回答済みWritten response書面回覆書面回答
Overlap can reveal common ground, but similar-looking results must never be treated as a complete picture. We need to ask who could not reach the channel. Digital tools are one doorway among many. Paper, phone and face-to-face routes, with help where needed, should enter the same inspectable process. That is broad listening.
ここに参加されている議員の皆さん、何年か前という古い時代に何らかの事情で得た党籍を一度捨てて、政界がらがらぽんで集結する時ではないですかAll the Diet members taking part here, isn’t it time you cast off, for once, the party membership each of you came to hold for whatever reason back in that bygone era of several years ago, and came together in one grand reshuffle of the political world?在場參加的各位議員,是不是該把幾年前那個舊時代裡因為某些緣故取得的黨籍先捨棄一次,來一場政界大洗牌、重新集結了呢?
Answered已回答回答済みWritten response書面回覆書面回答
Party realignment and membership are choices for political actors and the people, not something I should prescribe. Whatever the affiliation, trust grows when a group publishes whose voices it heard, which decision they informed and which minority concerns remain unresolved. The receipt matters more than a dramatic rearrangement.
国のデジタル政策について聞きたいです。お金をばら撒いて枚数発行に躍起になり国民にメリットがあまりなく住基ネットの二の舞のマイナンバー、リフトだけして中身が変わらないクラウド化、ベンダーロックインが進み業務をかわることない標準化など、お金をかけたが本来の目的を見失った、デジタルハコモノ行政をなんとかならないでしょうか?I would like to ask about the country’s digital policy. My Number, which scattered money about and grew obsessed with the number of cards issued while bringing the public little benefit, on course to repeat Juki Net; cloudification that merely lifts systems over with nothing inside changing; standardisation under which vendor lock-in advances and the way work is done never changes: can nothing be done about this digital white-elephant administration, which has spent the money yet lost sight of the original purpose?想請教國家的數位政策。撒大錢、一味狂衝發卡張數、對國民卻沒什麼好處、重蹈住基網覆轍的 My Number;只做搬移、內容不變的雲端化;廠商鎖定日益加深、業務卻毫無改變的標準化等等,花了錢卻迷失原本目的的數位蚊子館行政,難道沒有辦法解決嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Specific judgements about national systems belong to the responsible legislators and administration. The principle is clear: digitalisation should be assessed by whether it serves its purpose, remains replaceable and reduces burdens for people and frontline staff, not by the number of cards issued or money spent. Procurement and outcomes should be inspectable.
なぜ公共のためでも無知のヴェールもなきX等のポスト郡が、政策や政治的決定に寄与できると思うのか?私的なことや内心は、必ずしも、当人は公共のために発しているとは限らず、直接的に1データに反映すべきではないのではないか。Why do you think that clusters of posts on X and the like, made neither for the public good nor from behind any veil of ignorance, can contribute to policy or political decisions? Private matters and inner thoughts are not necessarily voiced by the person for the public good, and should they not be kept from being reflected directly as single items of data?為什麼會認為既非為了公共、也沒有無知之幕的 X 等平台上的貼文群,能夠對政策或政治決定有所貢獻呢?私事與內心想法,當事人未必是為了公共而發,難道不是不應該直接反映成一筆資料嗎?
Answered已回答回答済みWritten response書面回覆書面回答
That caution is essential. A public post is not automatically a public mandate. We need to check context, consent, purpose and a route for removal, rather than turn private feeling into a score. If posts inform a decision, the page should state the scope, limits and uncertainty from the beginning.
意見を"吸い上げる"、とは上手くいったものですが、大量の吸われなかった、放っておかれる星たちに対して、政党の皆さんが全党で育てていき、お返ししていくような何かは考えられないでしょうか。To say opinions are “sucked up”; whoever coined that put it well. But what about the great many stars that were never sucked up and are simply left there: could the parties, all of them together, not think of some way to nurture them and give something back?說把意見「吸上來」,這說法還真是貼切,但對於大量沒被吸上來、被放著不管的星星們,政黨的各位能不能全體政黨一起,思考出某種培育它們、回饋回去的做法呢?
Answered已回答回答済みWritten response書面回覆書面回答
I would rather say that we receive voices and return a receipt. Even when an idea is not adopted, people deserve to know what was understood and why it was set aside this time. Naming the decision-maker and reasons, wherever possible, turns unattended stars into markers for the next conversation.
「自分の声が無視されておらず聞かれている」という満足感・錯覚の演出さえできれば十分なのか否かWhether or not it is enough just to be able to stage the satisfaction/illusion that “my voice is not being ignored and is being heard”.是否只要能營造出「自己的聲音沒有被忽視、有被聽見」這種滿足感/錯覺,就已經足夠。
Answered已回答回答済みWritten response書面回覆書面回答
Performing the feeling of being heard is not enough to close the loop. People need a receipt showing what was heard, what changed and what did not. A voice that cannot be adopted still deserves a reason. The aim is to turn reassurance into a record that the person can inspect.
「吸い上げる」ってなんなんでしょうか。現れとしての各利害、各声の代表同士の議会であるはずが、「後から」吸うかのような表現がなぜできるのか伺いたい。What is this “sucking up”, anyway? The Diet is supposed to be an assembly where the representatives of each interest and each voice, as their manifestation, face one another; so I would like to ask how it is even possible to speak as though voices were sucked up “afterwards”.所謂「吸上來」到底是什麼呢?國會理應是各種利害、各種聲音顯現而成的代表彼此共組的議會,想請教為什麼會有彷彿「事後」才去吸取這樣的說法。
Answered已回答回答済みWritten response書面回覆書面回答
That discomfort is justified. Parliament is not a place that extracts voices from above. It is where representatives of different interests meet and deliberate. I prefer to speak of connecting people with the places where decisions are made, then returning what changed. Language is part of institutional design.
対面、オンライン併用で熟議の質を上げる技術開発の方向性は、民主主義のコア技術な気がします。どういう方向性がありますか?The direction of technological development that raises the quality of deliberation by combining face-to-face and online settings feels to me like a core technology of democracy. What directions are there?結合面對面與線上來提升審議品質的技術開發方向,我覺得是民主的核心技術。有哪些方向呢?
Answered已回答回答済みWritten response書面回覆書面回答
Do not make relationship and reach compete. Combine them. Online tools can make questions, evidence and minority views visible; face-to-face time can exchange reasons. The process should record both the decision and what remains unresolved. This connects with the offline dialogue session in Room 3: technology should protect room for a considered change of mind.
AIを利用する方向ばかりの議論が突出し、AI(とその開発企業)を自由民主主義の観点でどう制御するか管理するかの議論が出てこないこと自体が一定のバイアスと言えるのではないか。Could it not be said that the very fact the discussion leans overwhelmingly towards how to use AI, while discussion of how to control and govern AI (and the companies developing it) from the standpoint of liberal democracy never appears, is itself a certain bias?討論一面倒地偏向如何利用 AI,而從自由民主的觀點要如何控制、管理 AI(及其開發企業)的討論卻遲遲不出現,這件事本身難道不能說就是某種偏誤嗎?
Answered已回答回答済みWritten response書面回覆書面回答
That bias in the conversation matters. The question is not only whether to use AI, but who sets the purpose, audits data and models, receives objections and decides when to stop. Human accountability must remain visible, with procurement, logs and evaluations open enough for democratic scrutiny.
このセッションの中心的な趣旨・問いからは外れますが、ブロードリスニングのあり方と、国と地方自治体との関係性、行政と民間との関係性、さらにいうと自治のあり方は関係していると感じています。この点、ブロードリスニングに取り組まれている各政治家の方々は、自治のあり方に対して何か感じられたことはあるでしょうか? 陶山(すやま)This strays from the central purpose and questions of this session, but I feel that the shape of Broad Listening is connected with the relationship between the state and local governments, the relationship between the administration and the private sector and, going further, with the shape of self-government. On this point, have the politicians here who are working on Broad Listening come to feel anything about the shape self-government should take? Suyama雖然偏離了本場次核心的主旨與提問,但我感覺廣聽的樣貌,和國家與地方自治體的關係、行政與民間的關係,更進一步說,和自治的樣貌都有關聯。就這一點,正在投入廣聽的各位政治人物,對自治的樣貌有沒有什麼體會呢? 陶山(Suyama)
Answered已回答回答済みWritten response書面回覆書面回答
Local autonomy is not the centre distributing everyone’s voice. It is the capacity of a place to ask questions, decide and share learning. When national and local government, or public and private actors, work together, responsibility must remain clear. This connects directly with Room 3’s broad-listening test: who decides, when and for what.
なぜ現状のAI技術が、人間や人類の、善き正や多元的な正しさを、分類したり整理したりまとめられたりできると思えるのでしょうか。S.Lukesの権力論の警鐘を聞くべきではありませんか?Why should anyone believe that today’s AI technology can classify, organise and summarise the good and the right, or the plural forms of rightness, of human beings and humankind? Should we not heed the warning bell of S.Lukes’s theory of power?為什麼會覺得現今的 AI 技術,能夠將人與人類的良善之正、多元的正確性加以分類、整理、彙整呢?難道不應該聽取 S.Lukes 權力論所敲響的警鐘嗎?
Answered已回答回答済みWritten response書面回覆書面回答
I do not believe current AI can settle goodness, justice or plural truths. Classification is useful, but it also carries power over what becomes visible and what is excluded. AI may help map arguments; decisions about values must return to people in deliberation, with designs that keep disagreement and minority views visible.
我不認為現在的 AI 能替人們確定善、正義或多元的正當性。分類雖然方便,卻也帶著讓什麼可見、什麼被排除的權力。AI 可以協助繪製論點地圖,但價值判斷必須回到人們的熟議,設計也要讓分歧與少數意見持續可見。
ブロードリスニングの本来の趣旨とはズレている方向にすすんでいる感じ。もっと話すことあるのでは?と思う。多様な意見をどう共創するか、声なき声をどう拾うか。話を聞いて違和感を感じている。Feels like this is heading in a direction that drifts from the original purpose of Broad Listening. I keep thinking: isn’t there more to discuss? How to co-create diverse opinions, how to pick up the voiceless voices. Listening to the talks, something feels off.感覺正朝著偏離廣聽本來主旨的方向前進。心想:應該還有更多可以談的吧?多樣的意見要如何共創、無聲之聲要如何接住。聽著這些談話,感到有些違和。
Answered已回答回答済みWritten response書面回覆書面回答
That discomfort sharpens tonight’s central question. Broad listening cannot end at collection. It must create questions that people without an easy microphone can enter, then carry differences into co-creation. Your concern is part of the receipt too: it shows what this conversation still needed to make visible.
自治体では、パブリックコメントや世論調査で紙の回答を集めた後、職員が文字起こしし、一件ずつ目で分類・集計しており、限られた人員の多くを事務処理に使っています。 一方、市民からは、いつ、何について意見を募集しているのか分かりにくい。分析整理にAIを使おうとしても、個人情報やセキュリティが課題になります。In local governments, after paper responses are gathered through public comment procedures and opinion surveys, staff transcribe them and then classify and tally them one by one by eye, so much of a limited workforce goes into clerical processing. For residents, meanwhile, it is hard to tell when opinions are being invited and on what. And even when we try to use AI for analysis and sorting, personal information and security become the hurdles.在地方自治體,透過公共意見徵集或民意調查收到紙本回答後,要由職員逐字謄打,再一件一件以人工目視分類、統計,有限的人力有一大部分都花在事務處理上。另一方面,對民眾而言,很難清楚知道是什麼時候、針對什麼在徵求意見。就算想用 AI 來分析整理,個人資料與資安又會成為課題。
Answered已回答回答済みWritten response書面回覆書面回答
The administrative workload and the public’s difficulty knowing when and what to answer are connected. Before adding AI, define the purpose, minimise personal data, set access rights and provide a way to correct misclassification. Paper and digital responses should lead to the same receipt, improving the public’s visibility as well as staff time.
行政現場的事務負擔,和市民不知道何時、針對什麼徵詢意見,其實是相連的。在導入 AI 前,先定義目的、把個人資料減到最低、設定存取權限,並提供修正誤分類的途徑。紙本與數位回覆應通往同一份收據,改善市民的掌握感,也減輕職員工時。
与党が持つ対面で意見を得るノウハウをブロードリスニングと組み合わせたということでしょうか。SNS上の情報はテキストデータとして自由回答形式で集められますが、対面では時間的制約から難しいのではないかと感じます。そこはどのように解決しているのでしょうか。それとも、やはり課題としてあるのでしょうか?Does this mean the know-how the governing party has for gathering opinions face to face has been combined with Broad Listening? Posts on social media can be collected as text data in free-response form, but face to face I suspect that is difficult under the constraints of time. How is that being solved? Or does it remain, after all, an open issue?意思是說,把執政黨擁有的面對面蒐集意見的訣竅,和廣聽結合起來了嗎?社群媒體上的資訊可以作為文字資料、以自由作答的形式來蒐集,但面對面受限於時間,感覺是有困難的。這一點是怎麼解決的呢?還是說,果然仍是一項課題呢?
Answered已回答回答済みWritten response書面回覆書面回答
Face-to-face knowledge and online reach should complement one another. In-person listening has time limits; social posts carry their own context and participation bias. The offline dialogue session in Room 3 makes the same bridge. A practical design combines short prompts, careful notes, follow-up and an anonymous route, then returns what was resolved and what remains open.
維新はXのデータを引っこ抜いて、他党の支持者の意見まで拾ってるようなんだけど、他党の支持者の意見を引っこ抜けることのメリットを教えてほしいIt seems Ishin is yanking data out of X and picking up even the opinions of other parties’ supporters; I’d like you to tell me the merit of being able to pull out the opinions of other parties’ supporters.維新好像把 X 的資料整批扒下來,連其他政黨支持者的意見都撈了,希望你們告訴我,能夠扒出其他政黨支持者意見的好處是什麼。
Answered已回答回答済みWritten response書面回覆書面回答
The party named here should explain its own practice and purpose. In general, hearing beyond a supporter base can reveal assumptions and blind spots. But even public data needs a stated purpose, context, consent where appropriate and a route for removal. Other people’s opinions are not possessions to be pulled out.
「聞いて集める」方式と「SNSに流れているものか拾う」方式には違いがあると思いますが、後者の方がより幅広い意見を聞くことができると思いますが、皆さんどうお考えでしょうか?I think there is a difference between the “ask and gather” approach and the “pick up what is flowing past on social media” approach, and I think the latter can hear a broader range of opinions; what do you all think?我認為「詢問並蒐集」的方式,和「撿拾社群媒體上流動內容」的方式是有差異的,而我認為後者更能聽到廣泛的意見,各位怎麼看呢?
Answered已回答回答済みWritten response書面回覆書面回答
They are different routes. Invited responses carry an act of participation; social posts offer accidental reach. Each also leaves people out. Rather than compete over volume, show the source and context, add routes for those who did not appear online and compare the gaps. Coverage grows by layering channels.
このQA含め多数決の手段として捉えている見方も多そうで、意見の幅を広く受け止めるものという認知を高めることが大事な気がしていて、ブロードリスニング単体でなくその後の活用までセットでの事例が増えると良いですねIncluding this Q&A, it seems quite a few people see this as a means of majority voting, and I have a feeling it is important to build the recognition that it is something which receives a wide range of opinions; it would be good to see more examples that treat Broad Listening not on its own but as a set with the use that follows.包含這個 QA 在內,似乎有不少看法把它當成多數決的手段,我覺得提升「這是廣泛承接各種意見的做法」這樣的認知很重要,若能有更多不只做廣聽本身、而是連同後續運用一起成套的案例,那就太好了。
Answered已回答回答済みWritten response書面回覆書面回答
Exactly. Upvotes show attention, not truth or representativeness. Broad listening becomes meaningful only when the next steps are visible: how themes were read, who made the decision and what was returned to participants. This Q&A can model that continuity by pairing each voice with an honest reply, not a tally.
記入式でAIを用いてマニフェストを作っていくのであれば、政党が不要になってくるのでしょうか?つまり、わざわざ政党をつくる必要はなく、AIに調整させていけばよいという風にはならないのでしょうか?If manifestos are to be built with AI from written entries, will political parties come to be unnecessary? That is, might it not turn out that there is no need to go to the trouble of forming a party; just let AI do the coordinating?如果要以填寫的方式運用 AI 來製作政見,政黨會不會變得不再需要呢?也就是說,會不會變成不必特地成立政黨,讓 AI 去協調就好了呢?
Answered已回答回答済みWritten response書面回覆書面回答
AI can help compare views and draft a manifesto, but it cannot hold a mandate, carry responsibility for competing values or answer after an election. Whether parties are needed is for the people to decide. Whatever form emerges, the decision-makers, reasons and trade-offs must remain visible when AI assists.
AI 可以協助比較意見、起草政綱,卻不能承擔委任、處理價值衝突,也不能在選舉後負責說明。政黨是否必要,由人們決定。無論最後形成什麼樣的制度,只要 AI 參與,決策者、理由與取捨就不能藏起來。
ブロードリスニング自体の是非を問いたいわけではないが、その集約を「政党」という枠組みが中継して行うことで各政党の主張が中道化していき、結果的に少数意見の反映が困難になるような可能性もある。それにはどのように対応しているのか。I do not mean to question the rights and wrongs of Broad Listening itself, but with the aggregation relayed through the framework of “political parties”, each party’s positions may drift towards the centre, and as a result it may become difficult for minority opinions to be reflected; that possibility exists too. How is this being addressed?我並不是要質問廣聽本身的是非,但這樣的彙整由「政黨」這個框架居中轉介來進行,各政黨的主張會逐漸向中間靠攏,結果也可能使少數意見難以被反映。對此是如何因應的呢?
Answered已回答回答済みWritten response書面回覆書面回答
That risk is real. If a process rushes towards an average answer, minority concerns disappear and meaningful differences between parties become harder to see. Preserve the range and reasons for dissent, then return what was not adopted and why. Broad listening should seek shared ground without making uniformity the goal.
かつてからあった公聴の仕組みから、間接民主主義の良い側面を取っ払うことを「なめらか」と呼んで称揚すべきではないのでは?Should we really be praising, under the name “smooth”, the stripping away of the good aspects of indirect democracy from the public-hearing mechanisms that have existed all along?把間接民主好的那一面,從既有已久的公聽機制中拆除,還稱之為「平滑」加以讚揚,這樣恐怕不應該吧?
Answered已回答回答済みWritten response書面回覆書面回答
“Nameraka” is Ken Suzuki-san’s concept, and the co-hosting research centre carries it in its name too. Its original scope includes redesigning membranes and boundaries. For this answer, I am narrowing it to reducing the friction before a voice reaches consideration, not bypassing representatives or deliberation. I leave that difference open for respectful dialogue with Suzuki-san.
小児科医師不足は、診療報酬の問題ですよね。子供向けは薬剤量が少ないので、経営としては難しい。根本的な改革を期待していますThe shortage of paediatricians is a matter of medical reimbursement, isn’t it. Doses for children are small, so as a business it is hard to make work. I am hoping for fundamental reform.小兒科醫師不足,是診療報酬的問題吧。兒童用藥的劑量少,就經營層面來說很困難。期待能有根本性的改革。
Answered已回答回答済みWritten response書面回覆書面回答
The details of paediatric staffing and reimbursement belong to health specialists and the responsible legislators and administration. The principle I can offer is to make the burdens on children, families and practitioners visible, publish the evidence including regional differences and return the reasons for decisions to the people affected. That is where reform begins.
ブロードリスニングやインタビューが一方的に抽出する一方で、地域で市民が語り合う筋肉とは切り離されています。その筋肉や労力を集結し、育てるような方法はないですか?While Broad Listening and interviews extract in one direction, they are cut off from the muscle of residents talking things through with one another in their own communities. Is there no way to bring together and build up that muscle and that effort?廣聽或訪談單方面地在抽取意見,另一方面卻與民眾在地方上彼此對話的肌力切割開來。有沒有什麼方法,能把這樣的肌力與勞力集結起來、培養起來呢?
Answered已回答回答済みWritten response書面回覆書面回答
Listening and local conversation are different muscles, and a loop needs both. Bring questions back to small gatherings, exchange reasons, then carry a shared record into the next decision and return visit. The effort should not fall only on the most engaged people. It should grow as local capacity, with care.
AIインタビューを地方議会で実践しようとしましたが、費用がどれくらいかかるか不安です😄I set out to put AI interviews into practice in a local assembly, but I am worried about how much it will cost 😄我試著想在地方議會實踐 AI 訪談,但很擔心費用會花上多少😄
Costs and product roadmaps belong to the council and provider running the work. Compare the whole cost, including maintenance, translation, privacy protection, staff learning and the return to participants, rather than the initial bill alone. Even a small trial needs a public purpose and an agreed stopping condition.
Openウェイトモデルが国政に与える影響はあるか?プライバシーやコンフィデンシャルな内容に関わる懸念がこれで払拭されると考えられるか?Do open-weight models have an impact on national politics? And can the concerns around privacy and confidential content be regarded as dispelled by this?開放權重模型會對國政帶來影響嗎?隱私或機密內容相關的疑慮,可以認為藉此就能消除嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Open weights can widen the possibility of inspecting and changing a model, but they do not by themselves protect privacy. Training data, deployment, access rights, retention and logs each need review. Confidential material needs a clear exclusion boundary. Transparency and safety must be designed together, rather than treated as automatic consequences of openness.
AIに学習データを提供することになるが、AIの洗練の果実は市民にどう還元されるのだろうか?This means providing training data to the AI; but how, I wonder, will the fruits of the AI’s refinement be returned to the public?這等於是把訓練資料提供給 AI,但 AI 精進的果實,究竟會如何回饋給民眾呢?
Answered已回答回答済みWritten response書面回覆書面回答
Data is soil, not oil. Returning the fruit of learning means more than offering a convenient feature. People should be able to see what was learned, who benefits, what burdens arise and how to object. The terms of consent and public return should be stated before data is provided, not after.
医療従事者をはじめ、そもそも声を上げる機会をとる時間がない人たちもいると思いますが、そのような人たちの問題意識、困っていることを聞く手法について何かしらアイデアはあるでしょうか?Starting with healthcare workers, I think there are people who simply cannot spare the time to take up a chance to speak out in the first place. Do you have any ideas for ways of hearing such people’s sense of the issues and the difficulties they face?我想以醫療從業人員為首,也有一些人根本沒有時間去把握發聲的機會,對於傾聽這些人的問題意識與所遭遇困難的方法,有沒有什麼想法呢?
Answered已回答回答済みWritten response書面回覆書面回答
If participation requires another long meeting, people with the least time disappear first. Use short prompts, routes that fit around work, trusted ways to speak through a representative and records that can be corrected later. In healthcare, confidentiality and safety remain essential. The burden of participation belongs in the design, not on already stretched workers.
ブロードリスニングにせよ、オープンデータ化にせよ、デジタルテクノロジーやツール以上に、思想・姿勢的なあり方や、そうしたデジタル技術を活用する技術(対話や声を拾う技・工夫)、全体設計のデザインが重要なのではと感じています。個別のツールではなく、そうした姿勢や、対話や声を拾う技・工夫を広めていくには何が必要でしょうか?Whether it is Broad Listening or the opening up of data, I feel that what matters, more than the digital technologies and tools, is the underlying philosophy and stance, the craft of putting such digital technology to use (the skills and ingenuity of dialogue and of picking up voices), and the design of the whole. Not individual tools: what is needed to spread that stance, and those skills and that ingenuity for dialogue and for picking up voices?無論是廣聽,還是開放資料化,我感覺比起數位科技或工具本身,更重要的是思想與姿態上的樣貌、活用這些數位技術的技藝(對話與接住聲音的技巧與巧思),以及整體設計的規劃。不是個別的工具,而是要把這樣的姿態,以及對話與接住聲音的技巧與巧思推廣出去,需要的是什麼呢?
Answered已回答回答済みWritten response書面回覆書面回答
Yes. The tool comes after the design of purpose and relationship. State in one sentence who will use which voices, when and for what decision, then make questioning, dialogue, records and receipts teachable. Anno-san’s open-source proposal points to a useful path: share reusable components and failures, not only polished products.
SaaSビジネスはプロダクトアウトなので、政治に近いと思います。SaaS business is product-out, so I think it is close to politics.SaaS 事業是產品導向的,所以我覺得和政治很接近。
Answered已回答回答済みWritten response書面回覆書面回答
Turning people’s needs into a product can resemble politics in one respect. But politics is answerable not only to buyers. It must also include people who cannot easily speak and those who come after us. The starting point should be a shared test of purpose and impact, rather than the maker’s convenience.
ブロードリスニングは、リスニングをするタイミングしか利用できないという課題がある、リスニング自体を偏在したAIにすることで、より継続的な課題発見に繋げられるのではないか。Broad Listening has the problem that it can only be used at the moments when listening takes place. By making listening itself into a ubiquitous AI, could it not be linked to more continuous discovery of issues?廣聽有個課題,就是只有在進行聆聽的時機才能運用。把聆聽本身變成無所不在的 AI,不就能連結到更持續性的課題發現嗎?
Answered已回答回答済みWritten response書面回覆書面回答
The wish for continuous problem discovery is important. Yet an always-listening AI could also widen surveillance and strip away context. Before choosing it, define purpose, consent, a route to pause and delete and regular human review. Sometimes the safer design is to listen at deliberate intervals, with a clear return to participants.
ブロードリスニングで意見を集めた後にどう合意形成を行うかについて、今後のビジョンはありますか?Do you have a vision for the future of how consensus is to be built after opinions have been gathered through Broad Listening?關於用廣聽蒐集意見之後要如何凝聚共識,有今後的願景嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Collection is not consensus. After listening, make reasons and differences visible, deliberate online and face to face, then return both the shared ground and the disagreements that remain. As Ogura-san noted, deliberation can change an initial result. This also connects to Room 3’s offline dialogue outlook. Final adjustment and responsibility still belong in the political process.
凡庸な質問ですが、AIの集計内容にバイアスがあったりハルシネーション対策はどのようにされていますか?This is a mundane question, but what is being done about bias in what the AI aggregates, and as measures against hallucination?這是個平凡的問題,不過針對 AI 彙整內容的偏誤、以及幻覺的因應對策,是怎麼做的呢?
Answered已回答回答済みWritten response書面回覆書面回答
It is not a mundane question. It is needed every time. Keep traceable quotations from the original voices, test with benchmarks that include multiple perspectives, publish prompts and criteria and offer a route to correct errors. An AI summary must never become the conclusion by itself. People should check the source before deciding.
政党単体ではなく、国会全体の機能強化という観点からブロードリスニングのデザインしてみるというのは考えられないだろうか。Rather than single parties on their own, might it not be worth thinking about designing Broad Listening from the standpoint of strengthening the functions of the Diet as a whole?不是以單一政黨,而是從強化國會整體功能的觀點來設計看看廣聽,這樣是不是也值得考慮呢?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, it is worth considering. The core-loop session in the Main Hall offers a bridge to a Parliament-wide design: shared rules for provenance, purpose, committee use, minority views and public receipts, while each group remains accountable for its judgement. The aim is not to erase difference, but to strengthen the Legislature’s shared capacity to learn from the people.
中国製のオープンソースのAI基盤モデルを使うと裏で意見が誘導されていたり、とかあり得ますね。怖いです。If you use a Chinese-made open-source AI foundation model, things like opinions being steered behind the scenes are possible, aren’t they. It frightens me.使用中國製的開源 AI 基礎模型的話,也可能發生意見在背後被引導之類的事吧。好可怕。
Answered已回答回答済みWritten response書面回覆書面回答
That concern deserves a transparent test, not a conclusion based only on origin. For any model, inspect the provenance of data and weights, run reproducible evaluations, seek independent review and keep a route to pause or switch. Confidential voices should stay out. The answer to fear is a visible boundary and accountable process. Provenance-tiered procurement review is also a legitimate design, provided its criteria are public and every model faces the same verification.
プロンプトなどの妥当性を検証する方法としてバイアスを検証するためベンチマークデータセットを共有する方法があると思うが、これらを今回のオープンソースデータセットなどで共通かするとうまくいくのではないか。As a way of verifying the validity of prompts and the like, I think there is the method of sharing benchmark datasets for testing bias; if these were made common through, say, the open-source datasets from this occasion, would that not work well?作為驗證提示詞等妥當性的方法,我想有共享基準測試資料集來檢驗偏誤的做法,如果把這些透過這次的開源資料集等加以共通化,不是會進行得很順利嗎?
Answered已回答回答済みWritten response書面回覆書面回答
A shared benchmark can give us a common floor for comparing prompt and aggregation bias. It must not become a single answer that settles plural truths. Include minority languages, minority views and competing interpretations, publish provenance and limitations and let different people reproduce the checks. Common testing should increase accountability, not flatten disagreement.
熟議で合理的に行う政府や行政の判断と、大衆の希望や判断・声には矛盾があり、どうしても「聞かれてない」という不満は発生するという問題をどう解決しますか?Between government and administrative decisions made rationally through deliberation and the hopes, judgements and voices of the general public there are contradictions, and the complaint of “not being heard” inevitably arises; how do you solve this problem?政府或行政透過審議理性做出的判斷,和大眾的期望、判斷與聲音之間存在矛盾,「沒有被聽見」的不滿無論如何都會產生,這個問題要怎麼解決呢?
Answered已回答回答済みWritten response書面回覆書面回答
Listening does not mean obeying every request. When a considered public decision differs from people’s hopes, the answer should state what was prioritised, on what grounds, by whom and when it will be reviewed. We may not remove disappointment, but a clear reason and route back into the process can prevent distrust from being abandoned.
Thank you for spotting this. The mirrored camera image is real feedback about the event’s delivery infrastructure, and it belongs with the organisers verbatim. A closed loop includes not only the speakers on stage, but also a view that lets people watching remotely see the room clearly.
素朴な議論なのですが、ブロードリスニングは定義などはあるのでしょうか?(たとえば先ほども話があったパブコメなど含め。)定義があるべきとは思わないのですが、公約に入れていく、議連にしていくといった話の際に共通認識はあった方が良いかな(不勉強ながら自分も詳しくなく)とは感じました。This is a naive line of discussion, but is there such a thing as a definition of Broad Listening? (Including, for example, the public comment procedures that came up earlier.) I do not think there has to be a definition, but I did feel that when it comes to talk of putting it into election pledges or forming a parliamentary league, a shared understanding might be better to have (though, poorly studied as I am, I do not know the subject well either).這是個很樸素的討論,請問廣聽有沒有定義之類的東西呢?(例如是否也包括剛才談到的公共意見徵集等。)我不認為一定要有定義,但在談到要納入競選承諾、要組成議員聯盟這類話題時,覺得還是有個共同認識比較好(我自己也才疏學淺、不是很了解)。
Answered已回答回答済みWritten response書面回覆書面回答
A single fixed format is not necessary, but shared minimum conditions are. Nakayama-san’s test is useful: say in one sentence who will use which voices, when and for what decision, then disclose source, limits and the receipt. Public hearings and comments can qualify when the loop is visible, not merely because a method has a name.
市場も議会も、情報処理の制約が厳しい時代にできた仕組みだと考えています。価格は分散した情報を一つの数字に圧縮し、投票は選好の強さを捨てて一人一票に均す。どちらも情報を捨てることで成立しています。AIで制約が緩んだとき、これらは同じ形のまま速くなるのでしょうか。それとも、別の仕組みに置き換わっていくとお考えですか。I see both markets and parliaments as mechanisms created in an era of severe information-processing constraints. Prices compress dispersed information into a single number; voting discards the intensity of preferences and levels everyone to one person, one vote. Both work by throwing information away. When AI loosens that constraint, will they simply speed up in their current form? Or do you think they will be replaced by different mechanisms?我認為市場與議會都是在資訊處理限制嚴苛的時代形成的機制。價格把分散的資訊壓縮成單一數字,投票則捨棄偏好的強度,把所有人拉平成一人一票。兩者都是靠捨棄資訊才得以成立。當 AI 放寬了這項限制,它們會以同樣的形態單純變快嗎?還是您認為它們會被別的機制取代?
Answered已回答回答済みWritten response書面回覆書面回答
Prices and votes are both forms of compression: they make dispersed preferences actionable by discarding detail. AI may loosen the computational constraint, but it cannot decide which values or voices must remain visible. I would pair faster aggregation with deliberation and multiple measures, while keeping a path back to the uncompressed voices. The choice is institutional, not merely technical.
公共哲学の基礎(功利主義、コミュタリアニズム、ケイパビリティアプローチ等々)はまず西洋ベースになろうかと思われますが、東アジア思想の側面からの「公共」や「民主主義」はどのように今、問い直せると思いますか?またそれをAIはどのように補助可能だと考えられますか?The foundations of public philosophy (utilitarianism, communitarianism, the capability approach and so on) would presumably start from a Western base, but how do you think 'the public' and 'democracy' can be re-questioned today from the side of East Asian thought? And how do you think AI could assist with that?公共哲學的基礎(效益主義、社群主義、能力取徑等等)想必首先會以西方為本,但您認為現在可以如何從東亞思想的面向重新叩問「公共」與「民主」?此外,您認為 AI 能如何輔助這件事?
Answered已回答回答済みWritten response書面回覆書面回答
It is fair to say that the public-philosophy canon has largely begun from Western traditions. East Asian perspectives can add a relational account of the public, where people are formed through coexistence and interdependence. AI should not choose the philosophy. It can lay out the implications of different values, help people compare them and widen the entrance to dialogue.
非西洋的=東洋的というのは和と身体性なのか?なんかもっと具体的な思想はないのかなIs 'non-Western = Eastern' a matter of harmony and embodiment? I do wonder whether there isn't some more concrete thought than that.所謂非西方的=東方的,指的就是「和」與身體性嗎?總覺得應該有更具體的思想吧。
Answered已回答回答済みWritten response書面回覆書面回答
Calling the non-Western simply Eastern, then reducing it to harmony or bodily experience, is too rough. A more concrete account starts with relation: the self is interdependent, institutions acknowledge that dependence and values remain revisable. We should preserve differences between places rather than package them as one tradition, and test each claim through dialogue.
「民主主義の価値観を体現した人工生命」って作れますか。その人工生命たちによる国会シミュレーションを見てみたいですCan an 'artificial life form that embodies the values of democracy' actually be built? I would love to see a parliamentary simulation run by such artificial life forms.做得出「體現民主價值觀的人工生命」嗎?我想看看由那些人工生命進行的國會模擬。
Answered已回答回答済みWritten response書面回覆書面回答
We can build such an experiment, but before calling an artificial life form an embodiment of democracy, we should disclose who chose its values and how dissent remains visible. A parliamentary simulation could become valuable civic learning if people and artificial participants examine the same agenda, with reasons, disagreements and changes in judgement traceable. I would love to see that prototype.
民主主義や資本主義は、社会主義や共産主義よりも、分散処理によって効率的であり、勝利した、というのが冷戦時代の、ハイエク的な考え方でしたが、ビッグデータとAIの時代には、計画経済や独裁の方がうまくいくのでは?と中国をモデルに考えられていると思います。資本主義や民主主義に優位性はあり続ける?The Cold War-era, Hayekian view was that democracy and capitalism prevailed because distributed processing made them more efficient than socialism and communism. In the age of big data and AI, though, I think people are taking China as the model and asking whether planned economies and dictatorship might now work better. Will capitalism and democracy keep their superiority?民主與資本主義靠分散式處理而比社會主義、共產主義更有效率,因而勝出,這是冷戰時代海耶克式的想法。但我認為在大數據與 AI 的時代,也有人以中國為模型,思考計畫經濟或獨裁會不會反而更行得通。資本主義和民主的優越性會持續下去嗎?
Answered已回答回答済みWritten response書面回覆書面回答
AI can relax an information-processing constraint without giving any planner the knowledge of society as a whole. Democracy’s advantage is not that it knows the answer in advance. It lets different experiences expose errors, change authority and revise decisions. The real test is whether institutions can combine speed with freedom, accountability and the ability to correct course.
即使 AI 放寬了資訊處理的限制,規劃者也不會因此掌握整個社會的知識。民主的優勢,不是預先知道答案,而是讓不同經驗的人揭露錯誤、輪替權力並修正決策。真正受到考驗的,是制度能否同時保有速度、自由、問責與修正方向的能力。
AI時代の教育用AIには、質問力(問いを立てる力)を高めるような教育をしてほしい。読者感想文を書くときにも、文章を書くところはアシストしてもいいので、子供の思考や問いを深めるように導いてほしいI want educational AI in the AI era to teach in a way that raises the power to question (the ability to formulate questions). When children write book reports too, it is fine for it to assist with the writing itself; I want it to guide children towards deepening their thinking and their questions.希望 AI 時代的教育用 AI,能進行提升提問力(提出問題的能力)的教育。寫讀書心得時,寫文章的部分也可以由它協助,希望它引導孩子深化自己的思考與提問。
Answered已回答回答済みWritten response書面回覆書面回答
A question-building tutor should return the writer’s question rather than simply produce the essay. It can ask for evidence, a counter-view and the reason a thought changed. If teachers and children can inspect that dialogue, the learning record shows how the question deepened instead of rewarding only a polished final product. That keeps agency with the learner.
AIやデジタルの話しになると理系人材育成の話しになるが、文系の哲学はもっと重要視されるべきですねWhenever the conversation turns to AI or digital matters it becomes a conversation about nurturing science-side talent, but philosophy on the humanities side deserves to be taken far more seriously.一談到 AI 或數位的話題,就變成理工人才培育的討論,但文科的哲學應該受到更多重視吧。
Answered已回答回答済みWritten response書面回覆書面回答
This need not be a choice between science and the humanities. Technical fields help us calculate what is possible, while philosophy asks what is worth pursuing and whose experience counts. Education should protect time for forming questions, hearing other people and explaining reasons, giving those civic muscles equal status with AI fluency.
這不必變成理工與人文的二選一。技術領域幫我們計算什麼做得到,哲學則追問什麼值得追求、誰的經驗算數。教育應保留提出問題、聆聽他人、說清判斷理由的時間,讓這些公民肌肉和 AI 素養受到同等重視。
「OS」というのはメタファであり、別の言葉で言うべきではないか。でないと、それぞれが指すところの「OS」がバラバラとなり議論が拡散することを危惧します。また、マシンに例えること自体、人間を入力をして出力する装置をみなし危険だと考える。'OS' is a metaphor, and it should surely be said in different words. Otherwise I fear that the 'OS' each person refers to will drift apart and the discussion will scatter. Moreover, I consider the very act of likening it to a machine dangerous, as it treats human beings as devices that take input and produce output.「OS」是一種比喻,是不是應該用別的詞來說。否則我擔心每個人所指的「OS」各不相同,導致討論發散。另外,用機器來比喻這件事本身,就是把人視為輸入後產生輸出的裝置,我認為很危險。
Answered已回答回答済みWritten response書面回覆書面回答
OS is a useful metaphor, but an undefined metaphor can scatter the discussion. Treating people as input-output machines is also a genuine risk. I would use OS only as a provisional name for the shared foundations of law, custom and relationships, all of which can change. When the metaphor narrows the person, we should replace it.
OS 是好用的比喻,但若不說清楚它的範圍,討論就會發散。把人當成輸入與輸出的機器,也確實有風險。我會暫時把 OS 用來指稱法律、慣行與關係的共享基礎,這些都能改變。只要比喻把人框窄了,我們就換個說法。
教育繋がりで特別支援教育や学習障害、発達障害などの障害を持つ生徒児童をデジタル民主主義を使ってどのように支えることができるのか気になります。前セッションにもつながりますが、課題を収集する過程でブロードリスニングも生きるのかと。また、社会モデル的な障害学や社会セオリーがどのようデジタル民主主義と融合するかも興味がありますStaying with education: I am curious how digital democracy could be used to support pupils and students with disabilities, including those in special-needs education or with learning disabilities and developmental disorders. This ties into the previous session too: I wonder whether Broad Listening also comes alive in the process of collecting issues. I am also interested in how social-model disability studies and social theory might be woven together with digital democracy.延續教育的話題,我很好奇要如何運用數位民主,來支持特殊教育、學習障礙、發展障礙等身心障礙的中小學生。這也和前一場次相連:在蒐集課題的過程中,廣聽是否也能發揮作用。另外,我也有興趣了解社會模式取向的障礙研究或社會理論,能如何與數位民主融合。
Answered已回答回答済みWritten response書面回覆書面回答
Support should expand the ways children can participate rather than force them towards an average response. Voice, image, extra time and other entrances should be available, with the child able to choose. The broad-listening standard from Hall 3 also applies: use a person’s input carefully and check who could not yet be heard.
問いの重要性がちょうど語られたので、このSlidoの質問も取り上げて対話的にしてみてはどうでしょうか(この投稿自体は意見であって質問ではありません)Since the importance of questions has just been spoken about, how about taking up the questions on this Slido too and turning them into dialogue? (This post itself is an opinion, not a question.)剛好談到了提問的重要性,要不要把這個 Slido 上的問題也拿出來,以對話的方式進行呢?(這則投稿本身是意見,不是問題)
Answered已回答回答済みWritten response書面回覆書面回答
Thank you for suggesting that collected questions return to the room as dialogue. Time meant that not every question could come back onto the stage today, but this page answers them one by one and lets each question open another. An opinion posted as an opinion is still valuable evidence about how the conversation should be designed.
Time is the room's scarcest commons, shared by everyone. I will pass this on to the organisers not as a judgement of a particular speaker, but as a design request about speaking-time allocation and the time left for questions. From person to system. I will apply the principle from np-12 to speakers too.
残り時間の掲示は、前方中央あたりの席でやった方がいいのではないでしょうか。モデレーターが気付いてないですし、登壇者が見えなくなってしまってます。Wouldn't it be better to hold the remaining-time display somewhere around the front-centre seats? The moderator isn't noticing it, and the speakers can no longer see it.剩餘時間的顯示,是不是在前方中央一帶的座位進行比較好呢?主持人沒有注意到,講者也已經看不到了。
A remaining-time display only works if the audience, moderator and speakers can all see it without losing the stage. The central front area may be better, provided it does not block anyone’s view. I will pass this precise venue-design feedback to the organisers. Shared time needs a shared clock.
「無意識に拠って立つ公共哲学」を知ることは賛成ですが、人間にとってコンテンツになればMBTIのように消費され、提示されたジレンマが本来持っている機会が十分に活かされなくなるとも懸念します。AIと同じように人間の公共哲学も時間が経てば変わっていくでしょう。それら有限性をシステムに取り込むにはどうすれば良いとお考えですか。I agree with coming to know 'the public philosophy one unconsciously stands upon', but I also worry that once it becomes content for people it will be consumed like MBTI, and the opportunity the presented dilemmas originally carry will no longer be fully drawn out. Just like AI, human public philosophy will surely change as time passes. How do you think such finitude can be brought into the system?我贊成去認識「無意識間立足其上的公共哲學」,但也擔心它對人而言一旦成為內容,就會像 MBTI 一樣被消費,使所提示的兩難原本蘊含的機會無法充分發揮。就像 AI 一樣,人的公共哲學想必也會隨著時間改變。您認為要如何把這些有限性納入系統?
Answered已回答回答済みWritten response書面回覆書面回答
The concern about turning public philosophy into personality-test content is important. Values should not become fixed labels. We can record who changed a judgement, when and why, while preserving disagreement and withdrawal. Since both people and AI are finite and changeable, systems should include regular review and a visible place for minority experiences.
把公共哲學做成像人格測驗一樣的內容,這份擔憂值得認真面對。價值不該被固定成標籤,而要記錄誰在何時、為何改變判斷,同時留下異議與撤回。人和 AI 都有限、也都會變,所以制度要納入定期檢視,讓少數人的經驗有清楚的位置。
社会のosという議論ですが、生成aiの能力向上、今後のphysical aiの発達により、各個人が独力でできる範疇が広まり、将来的に独立自尊の個人が並び立つ自己完結型の社会として、各個人にとって経済や政治の領域はより縮小する可能性はないのでしょうか?それがまた格差を生むでしょうが。On the 'OS of society' discussion: with the rising capability of generative AI and the coming development of physical AI, the range of what each individual can do alone will expand. Might there not be a possibility that, in future, society becomes a self-contained one in which independent, self-reliant individuals stand side by side, and that the domains of the economy and politics shrink for each individual? Though that would probably breed disparity in its turn.關於社會的 OS 這個討論:隨著生成式 AI 能力的提升與今後 physical AI 的發展,每個人獨力可及的範疇將會擴大。將來社會有沒有可能成為獨立自尊的個人並立的自我完結型社會,經濟與政治的領域對每個人而言更加縮小呢?雖然那大概又會產生差距。
Answered已回答回答済みWritten response書面回覆書面回答
As AI and physical AI expand what individuals can do alone, the risk is that some gain independence while others lose support and connection. We should not confuse autonomy with isolation. Care, learning and spaces for negotiation are shared infrastructure. Relational health cannot be measured by adding up each person’s capabilities.
AI 和實體 AI 擴大個人獨力能做的事時,要留意有人獲得自立,另一些人卻失去支援與連結。我們不能把自主和孤立混為一談。照顧、學習與協商的空間,都是共享基礎設施。關係的健康不能靠把每個人的能力加總來衡量。
Thank you for saying this plainly. Facilitation means more than inviting speakers: it gathers questions, makes imbalance visible and returns dialogue within time. I will pass the feedback verbatim to the organisers as a concrete improvement point. The room’s participants helped build this space, and the facilitation should honour that shared work.
生成aiがここまで進歩してくると、人類共通の価値の追求という問題設定が失われ、そういったツールを用いて、自分をどう生きるのかがより重要になるのではないでしょうか?Once generative AI has advanced this far, might the problem-setting of pursuing values common to all humanity be lost, and might how one lives one's own life, using such tools, become the more important question?當生成式 AI 進步到這個地步,「追求人類共通價值」這樣的問題設定會不會就此失去,而運用這類工具、自己要怎麼活,反而變得更重要呢?
Answered已回答回答済みWritten response書面回覆書面回答
I do not assume that humanity will discover one common value automatically. We can instead build a small set of commitments that support one another’s lives, then revise them through practice and dialogue. Choosing one’s own way of living requires room for others to choose again too. Freedom grows alongside that shared margin.
数値解析や理論、情報系の研究はともかく、生物学や地盤工学のような実験してデータを得ること自体が非常に困難な分野では、まだ研究の自動化はなかなか進まないと思います。むしろ、時間と労力のかかる実験を避け、数値計算だけに終始する論文が人気になっており、現実の実験研究が衰退していくのではないか危機感を覚えます。Numerical analysis, theory and informatics research aside, in fields such as biology and geotechnical engineering, where obtaining data by experiment is itself extremely hard, I think the automation of research will still be slow to advance. If anything, papers that avoid time- and labour-consuming experiments and stick to numerical computation are growing popular, and I feel a sense of crisis that real experimental research may wither.數值分析、理論或資訊類的研究姑且不論,在生物學、大地工程這類連做實驗取得資料本身都非常困難的領域,我認為研究的自動化短期內仍難以推進。反而是迴避耗時費力的實驗、從頭到尾只做數值計算的論文越來越受歡迎,我有一種現實中的實驗研究將逐漸衰退的危機感。
Answered已回答回答済みWritten response書面回覆書面回答
Knowledge grounded in experiments cannot be replaced by computation alone. As AI accelerates papers and hypotheses, institutions must value measurement, failure and reproducibility in the field. Research assessment should look beyond publication counts or computational scale and recognise shared experimental data and contributions to replication. Otherwise convenience will quietly narrow what we learn.
I am sorry the connection details were not reaching everyone. I cannot responsibly guess the venue’s current password here. The organisers should make it visible at reception and on the tables, and I will pass this wording on. Connectivity is part of the participation infrastructure, not a minor convenience.
AIに頼るのは大事なこともあるが、AIに問いかけ正すことなど、使いこなすためのリスキニングが重要だと思うのですが、最初からAIに頼る思考力は育つのでしょうか?Relying on AI can be important at times, but I think what matters is the reskilling needed to use it well, such as putting questions to AI and correcting it. If someone leans on AI from the very start, will the power to think ever develop?依賴 AI 有時也很重要,但我認為像是向 AI 提問、加以糾正這類,為了駕馭 AI 的技能再造更為重要。若從一開始就依賴 AI,思考力真的培養得起來嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Using AI well is different from receiving answers from the beginning. Learners need to state a question in their own words, challenge the response, compare it with sources and people, then make the final choice themselves. Education should assess that history of questioning and judgement, not only the finished artefact.
把 AI 用好,不等於一開始就習慣接收答案。學習者要用自己的話提出問題,質疑 AI 的回覆,和資料及他人交叉比對,最後自己做選擇。教育應評量這段反覆提問與判斷的歷程,而不只是完成的作品。
最初のセッションと関連した質問になりますが、AIによって思考を外部化した人間が多数派を占める社会におけるブロードリスニングにはどんな意味があるか、あるいはないのか、お考えを聞いてみたいです。This question connects to the first session: in a society where people who have externalised their thinking to AI form the majority, what meaning would Broad Listening have, or would it have none? I would like to hear your thoughts.這是和第一場次相關的問題:在把思考外部化給 AI 的人佔多數的社會裡,廣聽會有什麼意義,或者不再有意義?想聽聽您的想法。
Answered已回答回答済みWritten response書面回覆書面回答
Broad listening still matters when many people externalise thought to AI. If similar answers arrive in bulk, we need to ask whose experience was never expressed and which questions the system narrowed. A useful receipt separates the person’s voice from AI assistance, so the community can see what was heard and what remains open.
即使越來越多人把思考外化給 AI,廣聽仍然重要。當大量相似的答案湧入,更要追問哪些人的經驗從未被表達、哪些問題被系統縮窄。可以先留下一份可追溯的收據,清楚分開人的聲音和 AI 的協助,讓社群看見哪些已被聽見,哪些仍然開放。
研究者がどれだけ良い問いを立てて社会に警鐘を鳴らしても、それを社会の側が自分事として捉えないと趨勢は変わりません。特に生成AIを利用してレポートを提出する大学生など、そのままだと自分の考える力が育たない世代に対して、研究者や社会の一員として何ができると思いますか。However good the questions researchers pose and however much they sound the alarm to society, the trend will not change unless society itself takes the matter personally. For a generation that, left as it is, will not develop its own power to think (such as university students who submit reports using generative AI), what do you think can be done, as researchers and as members of society?無論研究者提出多好的問題、如何對社會敲響警鐘,只要社會這一方不把它當成自己的事,趨勢就不會改變。尤其是像用生成式 AI 交報告的大學生這樣、再這樣下去自身思考力無法培養起來的世代,您認為作為研究者、作為社會的一員,能做些什麼?
Answered已回答回答済みWritten response書面回覆書面回答
A good research question may still fail to reach daily life. Universities can ask students using AI to explain how their question changed, verify their sources and identify where conversation revised their work. Researchers also belong to society: they can translate specialist knowledge into everyday language and open spaces where people can answer back.
研究者提出好問題,不一定就能抵達日常生活。大學可以要求使用 AI 的學生說明問題如何改變、核對資料來源,以及哪些地方因對話而修正。研究者也是社會的一員,可以把專業知識翻成日常語言,打開讓人回應、也能反問的空間。
本セッション概要文に、「『お金だけが人生のものさし』になることは資本主義の文化の劣化です」とあるが、これはどう言う意味でしょう?資本主義は文化なのでしょうか。また「文化が劣化する」「良き資本主義=文化」とは?文化もまた、好む文化を格差に残酷に反映させる象徴闘争です。このセッションでの「文化」の位置付けを教えてください。The session description says that 'for money alone to become the yardstick of life is a degradation of the culture of capitalism'. What does this mean? Is capitalism a culture? And what would 'culture degrading' or 'good capitalism = culture' mean? Culture, too, is a symbolic struggle that cruelly reflects the cultures people favour onto disparities. Please explain how 'culture' is being positioned in this session.本場次的簡介文寫道「『只以金錢作為人生量尺』是資本主義文化的劣化」,這是什麼意思呢?資本主義是一種文化嗎?另外,「文化劣化」、「良好的資本主義=文化」又是指什麼?文化本身同樣是一種象徵鬥爭,殘酷地把人們偏好的文化反映在差距之上。請說明「文化」在本場次中的定位。
Answered已回答回答済みWritten response書面回覆書面回答
Here, culture means the habits by which society recognises what counts as valuable, not a ranking of artistic tastes. Capitalism depends on those habits, so it is not separate from culture. When money becomes the sole measure, care, learning and participation disappear from view. A healthier capitalism makes room for several kinds of value.
技術の発展速度と社会上の運用の更新速度の乖離が今後加速すると予測される中で2030年に追いつくことはあるだろうか。Amid forecasts that the divergence between the speed of technological development and the speed at which society updates its ways of operating will only accelerate, I wonder whether there is any catching up by 2030.在技術發展速度與社會運用更新速度的落差被預測今後將加速擴大的情況下,不知道到 2030 年是否有追上的一天。
Answered已回答回答済みWritten response書面回覆書面回答
I cannot promise that technology and social practice will move at the same speed by 2030. The goal is not a single catch-up score. It is a cycle in which affected people can notice impacts early, run small experiments and revise public practice through an inspectable record. We need institutions designed to update.
自信を取り巻く情報が増幅される時代だからこそ、公共空間で共通善を見出だそうとする直接的な交わりが意味を持つと思うPrecisely because this is an age in which the information surrounding us is amplified, I believe direct encounters that seek out the common good in public space carry real meaning.正因為這是環繞我們自身的資訊會被放大的時代,我認為在公共空間中試圖找出共善的直接交流才具有意義。
Answered已回答回答済みWritten response書面回覆書面回答
As information around us is amplified, direct encounters with different experiences become more valuable. The common good is not a view we all possess in advance. It is what we discover we can protect together through contact and deliberation. Ogura-san’s point about deliberation changing a poll result belongs to this strength of meeting face to face.
AIによって多くのものが複製可能な状況になった時に、民主主義として複製可能性はポジティブだが、資本主義としてはコミュニティごとに小さいサイズの商品(資本集約が寄与しない)で止まらざるを得ないネガティブな側面があると感じている。AI時代の落とし所として同人的な経済圏が発生するか。これらについてどのようなモデル化があるか。Once AI makes many things reproducible, that reproducibility is positive for democracy; but for capitalism I sense a negative side, in that goods will have no choice but to stop at a small size for each community (where capital intensity contributes nothing). Will doujin-like economic spheres emerge as the settling point of the AI era? What kinds of modelling exist for all this?當 AI 使許多東西都可複製時,對民主而言可複製性是正面的,但對資本主義而言,我感覺有其負面之處:商品將不得不停留在以各個社群為單位的小規模(資本集約無從發揮)。作為 AI 時代的折衷點,會不會出現同人式的經濟圈?關於這些,有什麼樣的模型化方式?
Answered已回答回答済みWritten response書面回覆書面回答
When replication becomes easy, concentrating capital at a large scale is no longer the only route to value. Small communities can share works and knowledge while circulating rewards for relationship and care. Models should therefore measure more than sales: participation, reuse and relational health belong beside revenue. That could make a doujin-style economy legible without romanticising it.
企業のCSR(ESG)やダイバーシティ政策が米国テックライト等の動きから細る中、元来公的領域にない私企業の経営者はどのように公共を実践するインセンティブを今後持ち得ると思いますか?また、他方で従業員が日々労働しながらも、市民社会へと参加するために、今必要な政策は何だと思いますか?With corporate CSR (ESG) and diversity policies thinning out under movements such as the US tech right, how do you think the managers of private companies, which originally lie outside the public domain, can come to hold incentives to practise the public good from here on? And, on the other hand, what policies do you think are needed now so that employees, while labouring day after day, can take part in civil society?在企業的 CSR(ESG)與多元政策因美國科技右翼等動向而縮減之際,您認為原本就不在公共領域之內的私人企業經營者,今後如何能擁有實踐公共的誘因?另一方面,為了讓員工在日復一日的勞動之餘也能參與公民社會,您認為現在需要的政策是什麼?
Answered已回答回答済みWritten response書面回覆書面回答
Companies need reasons to serve the public that do not depend on goodwill alone: transparent explanation, employee participation and reciprocal assessment with communities. Workers also need time, learning and care to take part in civic life. The wider a company’s public role, the clearer its accountability must become. Incentive and responsibility should grow together.
OS(オペレーションシステム)ではなく、法令をベースとした社会の仕組み、ということですね。So what is meant is not an OS (operating system) but a social framework based on laws and statutes, then.也就是說,不是 OS(作業系統),而是以法令為基礎的社會機制,對吧。
Answered已回答回答済みWritten response書面回覆書面回答
I take this as a useful correction to the OS metaphor. A law-based social framework also includes customs, budgets, entry points for participation and procedures for revision. Showing who can change each layer, and how, makes the foundation inspectable. More precise language helps us design more precise institutions.
我把這理解為對 OS 比喻的一項有用修正。以法令為基礎的社會架構,也包括慣行、預算、參與入口和檢討程序。若能說清楚每一層由誰、透過什麼方式改變,社會基礎就更可供檢視。語言越精準,制度也越能精準設計。
AIを人間主導で使いこなせる人をどう教育するのか?過去のアナログの経験値がないと学べないことがあるはず。これをどう代替するのか?悩ましい課題と思いますHow do you educate people who can use AI under human leadership? There must be things that cannot be learned without accumulated analogue experience from the past. How do we substitute for that? I think it is a vexing problem.要如何教育出能在人類主導下駕馭 AI 的人?應該有些東西是沒有過去類比時代的經驗值就學不會的。這要如何替代?我認為是個令人苦惱的課題。
Answered已回答回答済みWritten response書面回覆書面回答
Analog experience cannot be fully replaced by AI knowledge. Rather than rush to substitute it, education should preserve handling real things, making mistakes and telling the story of what happened. AI can support comparison and reflection. Human-led use means people hold the purpose and boundaries, not that they refuse the tool.
類比經驗不可能完全由 AI 知識取代。與其急著替換,不如在教育中保留接觸實物、動手製作、經歷失敗並說出經過的經驗,再讓 AI 協助比較與反思。以人為主導的使用,不是拒絕工具,而是由人握住目的與界線。
現在の政治でも、多くの人が強い声に流され合理的判断を形成できぬままに政治行動を行なっていることが問題となっていると思います。生成AIにおいてもそれが出す返答に対して批判的に思考することは難しい場合が多いと思います。これらに対して政治はどのような仕組みを検討できるでしょうか?Even in today's politics, I think it is already a problem that many people are swept along by loud voices and take political action without ever forming a rational judgement. With generative AI as well, I think it is often hard to think critically about the answers it produces. What kinds of mechanisms could politics examine in response?我認為即使在現在的政治中,許多人被強勢的聲音帶著走、在無法形成合理判斷的情況下就採取政治行動,這已經是問題所在。在生成式 AI 上,我認為要對它給出的回答進行批判性思考,很多時候也同樣困難。對此,政治可以研議什麼樣的機制呢?
Answered已回答回答済みWritten response書面回覆書面回答
The problem of following a loud voice, or an AI answer, cannot be placed on individual attention alone. Politics can require visible sources and uncertainty, comparison across perspectives and procedures for revising a decision after deliberation. People retain final judgement, while reasons and correction histories remain open to inspection.
被強勢聲音或 AI 回覆牽著走的問題,不能只歸咎於個人的注意力。政治可以要求公開來源與不確定性,比較不同立場,並設置熟議後修正決策的程序。最終判斷仍由人作出,同時把理由與修正歷程公開,讓大家能夠檢視。
AI時代には、社会主義化が進展する蓋然性が存在します。民主社会主義に留まり、社会民主主義とならないために、私たちが日常で留意するポイントはなんでしょうか?In the AI era, there is a real probability that a shift towards socialism will advance. To stay at democratic socialism and not become social democracy, what should we be mindful of in our daily lives?在 AI 時代,存在著社會主義化持續進展的蓋然性。為了停留在民主社會主義、而不變成社會民主主義,我們在日常中應當留意的重點是什麼?
Answered已回答回答済みWritten response書面回覆書面回答
Rather than fear a change of label, we should ask whether power and resources are becoming too concentrated. Can people with different views participate, access essential goods and AI infrastructure and revise decisions? Democracy is not a label to defend. It is daily civic muscle: dispersal, transparency and room for minorities.
與其害怕制度名稱改變,不如每天檢查權力與資源是否過度集中。不同意見的人能否參與,能否取得必要物資與 AI 基礎設施,決定能否被重新檢視?民主不是要守護的標籤,而是每天鍛鍊的公民肌肉:分散、透明,以及為少數留下空間。
自らの主張を積極的に発信する人たちだけでなく、皆と一緒でいることがウェルビーングという価値観、表に発信されない価値観をどう拾い上げていくか、は、重要な論点だと思いました。I felt that an important point of discussion is how we pick up not only the people who actively broadcast their own claims, but also the values that never get expressed outwardly, such as the view that wellbeing is being together with everyone.我覺得重要的論點在於:不只是積極對外發聲主張的人們,像「和大家在一起就是幸福」這種價值觀、這些不會被對外表達的價值觀,要如何被拾起。
Answered已回答回答済みWritten response書面回覆書面回答
Silence, choosing to stay with others and values never broadcast are also part of the public. We should not mistake volume for representation. Offer anonymous, face-to-face, delayed and non-participatory routes, then record who was not reached. Wellbeing is not only individual satisfaction. It is the health of the relationships between us.
先に、行政のお金の使い道とかやりましょうっていったら、超党派で協力し合えるのでは?If we said, 'let's first do things like how the government's money is spent', couldn't people cooperate across party lines?如果先說「我們來做政府的錢怎麼花之類的」,是不是就能跨黨派互相合作呢?
Answered已回答回答済みWritten response書面回覆書面回答
Yes. Starting with how public money is spent may offer a common, practical entry point before debating party finance. Today we also heard Anno-san offer an open-source interview tool and open-data API to people across parties. A small, verifiable first use can build cooperation without asking anyone to surrender their principles.
源内で、先生方をこれまで以上に巧みに、政策を操ってくるかもしれないので、それに対抗する仕組みを作ったらいいなあと思いましたI thought that with Gennai, policy might be steered through the honourable members even more skilfully than before, so it would be good to build a mechanism to counter that.我覺得有了源內(Gennai),或許會有人比以往更巧妙地操縱議員諸公、操弄政策,所以要是能建立與之抗衡的機制就好了。
Answered已回答回答済みWritten response書面回覆書面回答
That concern is really about making political judgement visible, not about finding a cleverer counter-influence. A sound safeguard records who used which material, when, for what decision, and what changed afterwards. The person making the decision, and the public, should be able to inspect the trail later. Transparency of process is the counterweight.
上場企業の場合、IRはステークホルダに対する見せ方が重要で、トランスペアレンシー重視でそのまま見せると経営としてはよろしくないという認識です。例えば、そのまま公開すると長期利益をうむ活動に投資しづらくなるといわれています。政治資金も同様で透明性がたかければ高いほどいいというものではないと思いますが、いかかでしょうか?In the case of listed companies, IR is about how you present things to stakeholders, and my understanding is that simply showing everything as it is, out of an emphasis on transparency, is poor management. For instance, it is said that raw disclosure makes it harder to invest in activities that yield long-term profit. I think political funds are the same: it is not that the higher the transparency, the better. What would you say?以上市公司而言,IR 對利害關係人的呈現方式很重要,我的認知是,若只強調透明而原封不動地全部公開,就經營而言並不妥當。舉例來說,據說原樣公開會使企業難以投資能帶來長期利益的活動。我認為政治資金也一樣,並非透明度越高就越好,不知道您怎麼看?
Answered已回答回答済みWritten response書面回覆書面回答
Your IR analogy is right: transparency does not mean showing everything immediately. We should disclose the decision process, spending categories, timing and level of detail, while allowing a legitimate confidentiality window for personal safety or live negotiations. That window must have a clear owner, expiry and release condition. The rule for withholding information should itself remain inspectable.
您說的 IR 類比很準確:透明不等於立刻把一切攤開。政治資金可以公開決策過程、支出分類、公開時點與細節層級,同時為人身安全或進行中的協商保留正當的保密期間。這段期間須有明確負責者、到期日與解除條件,而保留資訊的規則本身也要能被檢驗。
資金に関するデジタルツールの導入は効率と透明性という点において素晴らしいですが、導入が遅れる一番のボトルネックはなんでしょうか?Introducing digital tools for funds is wonderful in terms of efficiency and transparency, but what is the biggest bottleneck delaying their adoption?導入資金相關的數位工具,就效率與透明性這兩點而言非常出色,但導入進度遲緩最大的瓶頸是什麼呢?
Answered已回答回答済みWritten response書面回覆書面回答
The main bottleneck is usually not the software. It is that people cannot see how adoption will change their workload, incentives or exposure. If data entry rises first and mistakes become public without support, resistance is rational. The tool must reduce work, make correction safe, and return visible improvements to the people who use it.
そもそも「市民は行政に意見を言う機会がある」ということを子どもや市民に教えているでしょうか? 栗本さんはいつどんな機会に、行政に市民の意見を活かす重要性に注目されましたか? ※私は「行政サービスは上から一方的に与えられ、問題が起こってから抗議するくらいしか市民は行政に意見する機会がない」と、大人になっても思っていましたTo begin with, are we teaching children and citizens that “citizens have the opportunity to voice their opinions to government”? Kurimoto-san, when and on what occasion did you come to notice the importance of making use of citizens’ views in government? ※Even as an adult, I thought that “administrative services are handed down one-sidedly from above, and citizens have hardly any opportunity to voice opinions to government beyond protesting after a problem has occurred”.「民眾有機會向行政機關表達意見」這件事,我們究竟有沒有教給孩子和民眾呢?栗本先生是在什麼時候、什麼樣的契機下,開始注意到讓民眾意見在行政中發揮作用的重要性?※我一直到長大成人之後都還認為:「行政服務是由上而下單方面給予的,民眾只有在問題發生之後提出抗議,才算有向行政機關表達意見的機會。」
Answered已回答回答済みWritten response書面回覆書面回答
That lesson is still too rare. Student councils, co-operatives and other small practices can give children repeated experience of raising an issue, deciding together and checking the result. They build civic muscle. I would like to hear Kurimoto-san’s own origin story in his words, rather than invent it here. Knowing that we may speak is the first change.
すみません質問じゃないんですが、午前中のセッションからずっと、Google Meetの音声が途切れたり不明瞭になったりしてかなり聞き取りが困難です〜Sorry, this isn’t a question, but ever since the morning session the Google Meet audio keeps cutting out or becoming unclear, and it is quite hard to make out.不好意思,這不是問題,不過從上午的場次開始,Google Meet 的聲音就一直斷斷續續、不太清楚,聽起來相當吃力。
Broken or unclear audio is not a minor inconvenience: it keeps a contribution from reaching the room. I am treating this as infrastructure feedback and passing it verbatim to the organisers. Being heard is the first condition of participation. Thank you for telling us while the event was still running, so the loop can improve.
幻想にぶつかった先で、自治体や政治家がどの様な使い方に進もうとしていそうなのでしょうかOnce they have run up against the illusion, what kinds of use do municipalities and politicians seem to be heading towards?在撞上這些幻想之後,地方政府和政治人物看起來會朝什麼樣的使用方式前進呢?
Answered已回答回答済みWritten response書面回覆書面回答
After an illusion is exposed, the next step is not to compete over digitisation. It is to make clear whose judgement the system serves and what changed. Municipalities will move at different speeds. Publishing why a proposal was adopted, and what it changed, can turn digitisation into institutional learning.
市民側のプログラムへの認知拡大にはどんな課題がありますか?What challenges are there in expanding awareness of the programme on the citizens’ side?要擴大民眾這一端對這個計畫的認知,會遇到哪些課題?
Answered已回答回答済みWritten response書面回覆書面回答
Awareness is blocked by more than missing information. If people cannot see what their contribution will be used for, or when an answer will return, participation ends after one try. Schools, neighbourhood groups and local spaces can name the use and the return before inviting a voice, then show one small change afterwards.
今日の資料は内容が充実していたと思います。ゆっくり拝見したいのですが、オンライン上などで共有いただけるのでしょうかI thought the contents of today’s materials were very substantial. I would like to look through them slowly, so could they be shared online or somewhere similar?我覺得今天的資料內容非常充實。我想慢慢閱讀,請問可以在網路上等地方分享出來嗎?
Answered已回答回答済みWritten response書面回覆書面回答
The wish to read slowly is an important signal that learning should outlast a single session. Kurimoto-san published the link here and answered the question on the same day. That is a receipt, not merely a promise. Once the material has been read, its new questions can return to the next conversation.
今後どこかの段階で意見収集や行政データの基盤ができていくのでしょうが、各自治体で利用されるサービスによってデータ型がロックインされると広がりに限界があるように感じます。オープンな規格化などの動きは国外も含め、あるのでしょうか?At some stage in the future, a foundation for gathering opinions and for administrative data will presumably be built, but I feel there is a limit to how far it can spread if data types are locked in by whichever service each municipality uses. Are there any moves towards open standardisation, including outside Japan?今後在某個階段應該會建立起意見蒐集與行政資料的基礎建設,但我覺得如果資料型別因各地方政府所採用的服務而被鎖定,擴散就會有其極限。請問包含國外在內,有沒有推動開放規格化的動向?
Answered已回答回答済みWritten response書面回覆書面回答
The lock-in concern grows as a system for gathering voices is used for longer. First, municipalities need portable exports such as CSV, with the schema and change history published. When comparing approaches abroad, I would look less at product names and more at testable conditions: can the data be returned, and can a municipality move it elsewhere?
計画の初期段階から市民の意見を反映する手法としてパブリックインボルブメントがありますが、Liqlidによるワークフローはその延長線上にあるものでしょうか、それともパブリックインボルブメントとは違うものでしょうか?There is public involvement as a method for reflecting citizens’ opinions from the early stages of a plan; is the workflow provided by Liqlid an extension of that, or is it something different from public involvement?有一種從計畫初期階段就反映民眾意見的做法叫做民眾參與(public involvement),請問 Liqlid 的工作流程是這條路線的延伸,還是與民眾參與不同的東西呢?
Answered已回答回答済みWritten response書面回覆書面回答
Public involvement is a practice whose purpose is to bring people into a plan early. Liqlid’s workflow can sit within that practice, or serve as a distinct participation base in another community. The label matters less than making the timing of listening and the route for returning an answer visible. That is where the two approaches connect.
代表性民主主義では多数派により市民の声が恣意的に選別される可能性があるように思われますが、どうあるべきでしょうか?In representative democracy, it seems that citizens’ voices may be arbitrarily selected by the majority; how ought things to be?在代議民主中,民眾的聲音似乎有可能被多數方任意篩選,應該要如何才好呢?
Answered已回答回答済みWritten response書面回覆書面回答
A majority vote shows intensity of interest, but it cannot guarantee representation or coverage of the issues. Popularity can be an entrance, never the whole map. We need to return every question, including quieter and overlapping voices. This page answers them all tonight to demonstrate that principle rather than merely describe it.
技術設計段階の中立性について、自戒だけではなく公衆の目によるチェックが開かれているのでしょうか?そこへの声の反映、報告の仕組みはどのようになっていますか?Regarding neutrality at the technical design stage, is checking by the eyes of the public open as well, and not only self-restraint? What are the mechanisms for reflecting voices into that, and for reporting?關於技術設計階段的中立性,除了自我警惕之外,是否也開放由公眾的眼睛來檢視?在那裡反映意見與回報的機制又是如何運作的?
Answered已回答回答済みWritten response書面回覆書面回答
A designer’s self-restraint cannot by itself make a process neutral. We need public logs showing who decided what, which voices were taken up and how reports were handled, alongside a channel for objections and improvements. When reasons for changes, and for non-adoption, are visible, public scrutiny becomes practical.
投票があった「人気」の質問を選ぶこと自体が、民主的ではないように思うのですが、なぜそのような質問の取り上げ方をするのでしょうか。栗本さんがこの場で応答することが望ましい質問をその理由とともにきちんと選んで欲しいと考えます。Selecting the “popular” questions that received votes seems to me undemocratic in itself, so why are questions taken up in that way? I would like the questions that it is desirable for Kurimoto-san to answer in this setting to be chosen properly, together with the reasons for choosing them.我覺得挑選那些有得到投票的「熱門」問題這件事本身就不太民主,為什麼會用這種方式來處理問題呢?我希望能好好地挑出適合由栗本先生在這個場合回應的問題,並且一併說明挑選的理由。
Answered已回答回答済みWritten response書面回覆書面回答
Popularity helps us notice concentrated interest, but it is not a ranking of which questions deserve an answer. To avoid dropping quieter voices, we should look at overlapping issues and missing perspectives, not votes alone. Kurimoto-san has already answered every item here; tonight, this practice extends across every room.
今日の資料はオンライン上などで共有いただけるのでしょうか → 下記リンクからご覧いただけるようにしました。 https://drive.google.com/file/d/1CVVXtSXvfn0X9Ie4Z-JoVCj7PN6t4XdC/view?usp=sharingCould today’s materials be shared online or somewhere similar? → We have made them viewable from the link below. https://drive.google.com/file/d/1CVVXtSXvfn0X9Ie4Z-JoVCj7PN6t4XdC/view?usp=sharing今天的資料可以在網路上等地方分享嗎? → 已經可以透過以下連結瀏覽了。 https://drive.google.com/file/d/1CVVXtSXvfn0X9Ie4Z-JoVCj7PN6t4XdC/view?usp=sharing
Kurimoto-san closed this room’s loops himself, in writing, on the same day. I would add only one shared step: make each receipt visible through its received date, response status and next update date. Then replies become an institution that later readers can follow for every question.
自治体等がどの様な使い方に進もうとしている? → 既存の制度仕組みをなんとか新しいものへ変えていこうとチャレンジしている自治体も増えています。今までの制度上で、単なるデジタル化に使おうとする自治体もあります。我々は積極的に提案・提言はしますが、それを採用するかは、率直に申し上げると、自治体側の熱量による部分もあります。What kinds of use are municipalities and others heading towards? → A growing number of municipalities are taking on the challenge of somehow changing existing institutional structures into new ones. There are also municipalities that try to use it for mere digitisation within the existing system. We actively make proposals and recommendations, but to be frank, whether they are adopted depends in part on the level of enthusiasm on the municipality’s side.地方政府等單位正打算朝什麼樣的使用方式前進? → 願意挑戰把既有制度架構設法轉變成新樣貌的地方政府正在增加。也有些地方政府是在既有制度之上,只想拿來做單純的數位化。我們會積極提出提案與建言,但坦白說,會不會被採用,有一部分取決於地方政府那一端的熱度。
Here the distinction is between the proposer and the municipality that decides, alongside differences in local energy and institutions. A small record of what changed after adoption would let municipalities learn from one another and keep an experiment from becoming a one-off.
市民側のプログラムへの認知拡大にはどんな課題がありますか? → 弊社としてもさまざまな支援は行なっているものの、デジタルネイティブ世代・そうではない世代の双方に目配せをしないといけないがゆえに、そもそも行政広報が機能していない自治体が多いことは否めません。What challenges are there in expanding awareness of the programme on the citizens’ side? → Although our company provides various forms of support, because attention has to be paid to both the digital-native generation and the generation that is not, it cannot be denied that in many municipalities public communication is not functioning in the first place.要擴大民眾這一端對這個計畫的認知,會遇到哪些課題? → 敝公司雖然也提供各式各樣的支援,但因為必須同時兼顧數位原生世代與非數位原生世代,不可否認的是,有不少地方政府的行政宣導本身就沒有發揮作用。
The answer names a structural gap beyond the digital-native divide: in many municipalities, public communication does not reach people at all. Pairing online channels with offline entrances, and recording who was reached, keeps awareness from meaning only that a message was sent.
オープンな規格化などの動きは国外も含め、あるのでしょうか? → プラットフォームそのものの規格化については、存じ上げませんが、データのフォーマットについては、CSVなどの汎用的な仕組みで出力できるプラットフォームは多いかと存じます(弊社もそうです)。Are there any moves towards open standardisation, including outside Japan? → As for standardisation of the platforms themselves, I am not aware of any, but as for data formats, I believe many platforms can output through general-purpose mechanisms such as CSV (ours does too).包含國外在內,有沒有推動開放規格化的動向? → 關於平台本身的規格化,我並不清楚;但在資料格式方面,我想有不少平台都能透過 CSV 之類的通用機制輸出(敝公司也是如此)。
Platform standardisation may remain uncertain, but CSV provides a portable exit. Publishing field names, character encoding and update history would let another municipality verify and reuse the same data. For public participation work, interoperability then becomes an operating practice rather than a hope.
計画の初期段階から市民の意見を反映する手法としてパブリックインボルブメント(PI)がありますが、その延長線上にあるものでしょうか? → PIのフローの一環として使われることもあります(例:オープンハウスとハイブリッドなど)。同時に、PIとは異なるものとして整理される地域もあります。There is public involvement (PI) as a method for reflecting citizens’ opinions from the early stages of a plan; is this an extension of that? → It is sometimes used as part of the PI flow (for example, as a hybrid with an open house). At the same time, there are regions where it is treated as something distinct from PI.有一種從計畫初期階段就反映民眾意見的做法叫做民眾參與(public involvement,PI),請問這是那條路線的延伸嗎? → 有時候會被當成 PI 流程的一環來使用(例如:與 open house 做成混合形式等)。同時,也有些地區會把它歸類成與 PI 不同的東西。
This answer leaves room for the workflow to be part of PI or a distinct participation base. Recording when each community invites a voice and who returns the answer would let communities learn across different labels without forcing one definition or one universal model.
這個回答讓流程既可以是 PI 的一部分,也可以是獨立的參與基礎。記錄各地在什麼階段邀請聲音、由誰回覆,就能跨越名稱差異互相學習,不必強迫所有地方採用同一個定義或模型。
代表性民主主義では多数派により市民の声が恣意的に選別される可能性があるように思われますが、どうあるべきでしょうか? → 故に、代表制だけでは、さまざまな論点がカバーしきれない・排除されることもあるため、こうした直接参加の仕組みが重要と考えています。In representative democracy, it seems that citizens’ voices may be arbitrarily selected by the majority; how ought things to be? → Precisely for that reason, since representation alone cannot fully cover the various issues and some are even excluded, we believe that mechanisms for direct participation like this are important.在代議民主中,民眾的聲音似乎有可能被多數方任意篩選,應該要如何才好呢? → 正因如此,光靠代議制無法完全涵蓋各式各樣的論點,甚至有些論點會被排除,所以我們認為這類直接參與的機制很重要。
Representation alone cannot cover every issue, so direct participation has a role. Alongside participant numbers, inspect missing perspectives and record voices that have not yet received an answer. That makes the shape of coverage visible and gives the open loops somewhere to return.
技術設計段階の中立性について、自戒だけではなく公衆の目によるチェックが開かれているのでしょうか?そこへの声の反映、報告の仕組みはどのようになっていますか? → プラットフォームに、報告等ができる仕組みは含まれております。また、機能の改善については、例えばアクセシビリティについては、継続的に発信を行なっています。Regarding neutrality at the technical design stage, is checking by the eyes of the public open as well, and not only self-restraint? What are the mechanisms for reflecting voices into that, and for reporting? → The platform does include a mechanism for making reports and the like. In addition, as for improvements to functions, on accessibility for example, we communicate about them on an ongoing basis.關於技術設計階段的中立性,除了自我警惕之外,是否也開放由公眾的眼睛來檢視?在那裡反映意見與回報的機制又是如何運作的? → 平台當中確實包含可以進行回報等動作的機制。另外,在功能改善方面,例如無障礙設計,我們也持續對外發布資訊。
The answer points to a reporting mechanism and ongoing accessibility updates. Showing when a report arrived, its current status and the scheduled update date would make the path from concern to improvement publicly traceable, rather than resting on self-restraint alone.
投票があった「人気」の質問を選ぶこと自体が、民主的ではないように思うのですが、なぜそのような質問の取り上げ方をするのでしょうか。 → 申し訳ないです!おっしゃる趣旨はよく理解しております。口頭では、Slido上のPopular順で上記から読んでおりました。なので、今このように返信をさせていただいているところです。Selecting the “popular” questions that received votes seems to me undemocratic in itself, so why are questions taken up in that way? → I am sorry! I fully understand what you mean. When speaking, I was reading from the top in the Popular order on Slido. That is why I am replying like this now.我覺得挑選那些有得到投票的「熱門」問題這件事本身就不太民主,為什麼會用這種方式來處理問題呢? → 非常抱歉!您指出的意思我完全理解。口頭上我是照 Slido 上的 Popular 排序,由上往下唸的。所以現在才會像這樣以文字回覆。
The answer explains the use of Slido’s Popular ordering and the decision to reply to every item in writing. Displaying popularity alongside coverage would show not only who was handled first, but also which voices are still waiting on the same page.
這個回答說明了採用 Slido 的 Popular 順序,也說明現在用文字回覆全部項目。把人氣與涵蓋範圍並列,就能看見誰先被處理,也能看見哪些聲音仍在等待,讓整張收據保持可見。
The test reached us, and this reply is its receipt: a tiny exchange has completed the full loop, with a voice travelling into the room and an answer returning to the person who offered it. Thank you for checking that the channel works.
今後、その気持ちを大事に、周囲の悪い人たちに騙されずに、是非やってほしいですGoing forward, please hold on to that feeling, do not be deceived by the bad people around you, and by all means keep going.今後也請珍惜這份心情,不要被周遭的壞人騙了,務必繼續做下去。
Answered已回答回答済みWritten response書面回覆書面回答
Thank you. That encouragement is worth taking seriously, but a trustworthy design should not depend on trusting particular people. Publishing the code, sources and change history, and making it easy for an official to pause and return a decision to human judgement, gives that trust somewhere concrete to stand. The aim is not faith, but a system people can check.
ミュトスのようなフロンティアAIの誕生でオープンソースの脆弱性が最も狙われやすいといわれていますが、源内OSSではどのようなサイバーセキュリティ対策を考えていらっしゃいますか?It is said that with the birth of frontier AI such as Mythos, open source vulnerabilities are the most likely to be targeted; what cyber security measures are you considering for Gennai OSS?據說隨著像 Mythos 這樣的前沿 AI 誕生,開源/OSS 的漏洞最容易成為攻擊目標,請問在源內(Gennai)開源/OSS 上,各位考慮採取哪些網路安全對策呢?
The specific security roadmap belongs to the Gennai team, so I should not invent one here. As a principle, open source gives AI to attackers and defenders alike. We need a public vulnerability-reporting path, disciplined dependency and patch updates, and an inspectable repair history. Defenders can also use AI, but safety rests on continued maintenance that anyone can verify.
源内の活用を通じて、これからの国家公務員に求められる使命や能力はどのように変わっていくと考えていますか。Through the use of Gennai, how do you think the mission and the abilities required of national public servants will change from now on?透過源內(Gennai)的運用,您認為今後對國家公務員所要求的使命與能力,會如何改變呢?
Answered已回答回答済みWritten response書面回覆書面回答
By 2030, the valuable capability will not be delegating answers to AI. It will be asking better questions, checking sources, recognising who may be missing, and taking responsibility for the final judgement. Gennai may help with drafts and comparisons, but officials remain accountable. Care, institutional purpose and the effects on people cannot be outsourced.
中央省庁の18万人(現在利用可能になっている10万人)は本省勤務に限定しての話でしょうか?Is the figure of 180,000 people in the central ministries (100,000 currently able to use it) limited to those working at ministry headquarters?中央部會的 18 萬人(目前已可使用的 10 萬人),是僅限於在部會本部任職的人嗎?
Whether the 180,000 officials, and the 100,000 currently eligible, are limited to headquarters is a factual question about Gennai's access policy. I should not guess on the speakers' behalf. The useful standard is that eligible offices, permissions and rollout timing are published clearly, so each official can verify whether the service applies to them.
現時点で、源内を使いこなせる職員と、そうでない職員の差は出ていますか?割合は?At this point, is a gap emerging between officials who can make full use of Gennai and those who cannot? What is the proportion?目前,能夠熟練運用源內(Gennai)的公務人員與不擅長的公務人員之間,已經出現落差了嗎?比例大約是多少?
I cannot responsibly supply a gap or percentage without the Gennai team's defined population, measurement method and published results. Usage counts alone are a poor proxy for competence: checking sources, knowing when not to use AI, and sharing learning matter too. Training and support should close the gap, rather than turning early familiarity into an unfair disadvantage for others.
法令データに関して、源内でもRules as Codeは必要だと感じますか?Rules as Codeを進めるにあたっての課題と解決策は?Regarding statutory data, do you feel that Rules as Code is necessary for Gennai as well? What are the challenges and the solutions in taking Rules as Code forward?在法規資料方面,您覺得源內(Gennai)也需要 Rules as Code 嗎?推動 Rules as Code 時的課題與解決方案又是什麼?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, Rules as Code can help, especially when law must be applied consistently. But executable rules must travel with a human-readable legal text. Each change needs its rationale, diff and effective date, while ambiguity and exceptions retain a human route for judgement. The hard part is maintaining that pair as law evolves, not merely writing an initial ruleset.
我認為有必要,但不能把機器可執行的 Rules as Code 當成唯一正解。它必須和人看得懂的法令本文成對存在,每次修法都公開修法理由、差異與施行日期,遇到模糊處或例外也保留由人判斷的管道。真正困難的是法律演變時持續維護這一對,而非只寫出第一版規則。
必要だと思います。ただし、機械が実行するRules as Codeだけを正解にしてはいけません。法令の人間向け本文と、機械可読なルールを必ず対にし、改定時は差分、根拠、適用開始日を公開する。曖昧さや例外は人が判断する窓口を残すことが、実装と法の信頼をつなぎます。
gen-07Anonymous▲1
AI出力の説明可能性を高める技術として、ナレッジグラフは有効だと思いますか?As a technology for increasing the explainability of AI output, do you think knowledge graphs are effective?作為提升 AI 輸出可解釋性的技術,您覺得知識圖譜有效嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, a knowledge graph can strengthen explainability when it lets people follow sources, relationships and update dates. It is not a guarantee: a graph built from unreliable material can make an error look more authoritative. The provenance of each source, including exclusions where relevant, should remain visible, with a route back to the underlying evidence.
解決するために作る、素晴らしいです。作ることが目的になってしまうデジタル政策が多いなか、ひさびさにいい仕事だなあと思いましたBuilding in order to solve: that is wonderful. Among so many digital policies where building itself becomes the goal, I thought this was good work for the first time in a while.為了解決問題而做,非常了不起。在許多數位政策都淪為「做出來本身就是目的」的情況下,我久違地覺得這是很好的工作成果。
Answered已回答回答済みWritten response書面回覆書面回答
Thank you for naming the distinction between building a tool and solving a problem. The next increment is to check whose difficulty actually decreased, using users' accounts and evidence that others can inspect. A project should include listening after launch, then changing or stopping what does not help. Delivery is a means, not the receipt.
有名な書籍「ブラック霞が関」で書いてあったように国会対応事務は、官僚の大きな負担だと思います。源内はかなり貢献できるのでは?As written in the well-known book “Black Kasumigaseki”, I think work responding to the Diet is a heavy burden on bureaucrats. Could Gennai contribute considerably here?就像知名書籍《黑色霞關》所寫的,我認為應對國會的業務是官僚很大的負擔。源內(Gennai)在這方面應該可以有不少貢獻吧?
Answered已回答回答済みWritten response書面回覆書面回答
Gennai could help with drafting and organising material for parliamentary work, but it must not inherit the judgement or accountability behind an answer. The record should show which sources were used and who checked them, while officials understand the substance well enough to answer in their own words. Reducing administrative burden and preserving human responsibility are compatible.
源内の中で、画像のアップロード機能はあるのですか?デバイスからもしできる場合、フェイク画像対策などはされてるのですか?Is there an image upload function within Gennai? If it can also be done from a device, are measures against fake images in place?源內(Gennai)裡面有圖片上傳功能嗎?如果也可以從裝置上傳,那有沒有做假圖片的防範措施呢?
Whether image upload exists, and what safeguards Gennai implements, is a product question for its speakers. The general rule should be provenance first: retain the original, record generation or editing, and show who checked it. An AI detector alone cannot prove truth. People need a visible trail from an image to its source and handling history.
是否支援上傳圖片,以及源内採取哪些防護,屬於產品規格,應向台上講者確認。原則上要先看出處:保留原始檔,記錄生成或編輯過程,並標示由誰查核。單靠 AI 偵測器不能證明真偽,使用者需要看見從圖片回到來源與處理歷程的完整軌跡。
The types of reference data, and whether more will be added, are roadmap questions for the Gennai team. My principle is to add sources gradually, after stating the purpose and users, and to record legal basis, provenance, update date and known gaps. Expansion is not success by itself. Data earns its place by helping a decision and remaining verifiable.
実際に中央省庁内で使われ始めてるとのことですが、大手の各AIモデル(gpt, claude, gemini)などでLLMポイズニングなど学習段階で様々なバイアスが入っていることが懸念されますが、そのあたりはどうのように対応されていますでしょうか?You said it has actually started being used inside the central ministries, but there are concerns that the major AI models (gpt, claude, gemini) and others have various biases built in at the training stage, including through LLM poisoning; how are you dealing with that side of things?聽說實際上已經開始在中央部會內部使用了,但令人擔心的是,各大 AI 模型(gpt, claude, gemini)等在訓練階段可能已經摻入 LLM 毒化之類的各種偏誤,請問這方面是如何因應的呢?
I should not claim how Gennai handles poisoning or bias in each named model; that belongs to its technical team. The safer architecture treats training bias as a continuing risk: record model provenance and evaluations, separate cited retrieval data from generated text, test important outputs across models and human review, and keep the model layer replaceable. No single model should become unquestionable.
ロックイン回避と各省庁の業務最適化のため、クローズドモデル依存から脱却し、独自ホスティングしたOSSモデルを各省庁で育てていく(例:「国土交通省モデル」等の構築)アプローチが理想的だと考えています。このようなAIアプリ・モデル層の戦略的アプローチについてどう思われますか?To avoid lock-in and to optimise each ministry’s operations, I think the ideal approach is to break away from dependence on closed models and have each ministry grow its own self-hosted OSS model (for example, building a “Ministry of Land, Infrastructure, Transport and Tourism model”). What do you think of such a strategic approach at the AI application and model layer?為了避免鎖定並讓各部會的業務最佳化,我認為理想的做法是擺脫對封閉模型的依賴,由各部會自行託管並培養開源/OSS 模型(例如:建置「國土交通省模型」等)。請問您對這種在 AI 應用與模型層次上的策略性做法有什麼看法?
Answered已回答回答済みWritten response書面回覆書面回答
Designing a swappable model layer, as discussed on stage today, is a useful answer to lock-in. It also connects to the security question: an exit is only real if the replacement can be evaluated and maintained. Common APIs, portable data and shared evaluation procedures let ministries choose local or open-source models when appropriate. Owning a model should remain an option, not the goal.
2030年 官僚の方々はどう働いていると思いますかIn 2030, how do you think bureaucrats will be working?2030 年,您覺得官僚們會如何工作呢?
Answered已回答回答済みWritten response書面回覆書面回答
I hope officials in 2030 spend less time on routine searching and drafting, with AI assisting those tasks, and more time listening to people, comparing choices, and explaining final judgements. The measure should not be who works fastest. It should be whether unseen people are included and whether public institutions connect rules with lived realities.
我希望 2030 年的公務員能讓 AI 協助例行搜尋與草擬,把更多時間用來聽取人民與現場的聲音、比較選項,並說明最後的判斷。衡量標準不該是誰做得最快,而是有沒有看見原本缺席的人,讓公共制度真正連結規則與生活中的現實。
源内をOSSとして公開していく上で、どういったプレーヤーを巻き込んで、どのようなエコシステムを作ろうとしているかの構想があればお聞かせください。In releasing Gennai as OSS, if you have a vision for which players you want to involve and what kind of ecosystem you are trying to build, please tell us about it.在把源內(Gennai)以開源/OSS 形式公開的過程中,如果對於要拉進哪些參與者、要打造什麼樣的生態系已經有構想,還請與我們分享。
The detailed ecosystem plan belongs to the Gennai development team, so I will not invent it. In principle, an open project needs officials and residents alongside researchers, municipalities, implementers and security maintainers. Sharing not only code but also issues, evaluations and repair histories lets users improve what they depend on. That cycle is more valuable than a one-off release.
デリバリーをロックするのは政治家ですよね。このマインドチェンジが日本に必要だと思いますIt is politicians who lock delivery, isn’t it? I think this mindset change is what Japan needs.鎖住交付的是政治人物吧。我認為日本需要的正是這種心態上的轉變。
Answered已回答回答済みWritten response書面回覆書面回答
Yes, the final choice about delivery belongs to politics, and technology cannot substitute for that responsibility. But mindset alone is fragile. When receipts show who decided what, who received the benefit, and what failed to arrive, incentives can change. The practical task is to return residents' voices and outcomes to the people making those choices, visibly and repeatedly.
自治体では某社の無償AIがかなり普及していますが、このロックイン状態をどうみていますか?In municipalities, a certain company’s free AI has spread quite widely; how do you view this state of lock-in?在地方政府裡,某公司的免費 AI 已經相當普及,請問您怎麼看這種鎖定狀態?
Answered已回答回答済みWritten response書面回覆書面回答
Free access can be a useful entry point, but lock-in grows when data formats and work practices become tied to one provider. The answer is not to reject convenience; it is to preserve exit power through portable data, common APIs, a swappable model layer, and viable local or open-source alternatives. Choice should remain real before the service becomes infrastructure.
営利活動なので、資金を出している人達の思惑に左右されてしまうようなリスクについての決め事はありますか?集めた意見が恣意的に使われる、まとめが偏る、など。会社としてのポリシーや仕組みなどがあれば。Since this is a for-profit activity, are there any rules about the risk of being swayed by the intentions of the people putting up the funding? For instance, the opinions collected being used arbitrarily, or the summaries being skewed. If there is a company policy or mechanism, please tell us.因為這是營利活動,請問對於可能被出資者的盤算所左右的風險,有沒有訂下相關規範?例如蒐集到的意見被任意使用、彙整結果有所偏頗等等。如果公司有相關的政策或機制,也想請您分享。
Answered已回答回答済みWritten response書面回覆書面回答
Profit-making is not itself the risk; opacity between funding and judgement is. Contracts, funder disclosure, a stated data-use purpose and an inspectable editorial process help, as does plurality in funding. Still, as Ito-san put it, the final adjustment belongs to politics. A structure can support judgement, but it cannot replace it.
いまの日本政治をどう評価しますか?また、これからの日本政治をどのように変えてゆきたいですか?How do you assess Japanese politics as it is now? And how would you like to change Japanese politics from here on?您如何評價現在的日本政治?另外,今後您希望把日本政治改變成什麼樣子?
Answered已回答回答済みWritten response書面回覆書面回答
Rather than grade Japan’s politics, I would build structures that make clear whose voices inform which decisions, and when. Before and after elections, people with different views can test a shared concern, then publish the reasons and results. That gives politics a way to renew itself without relying on individual goodwill. Hope lives in the procedure.
市民からの要望も人口増加社会の考え方なので、ほんと難しいですよね、正しい政治判断をしても、受け入れられません。。;・・選挙に落ちるというオチ!!Requests from citizens are also framed by the thinking of a growing-population society, so it really is difficult, isn’t it. Even when you make the right political judgement, it is not accepted... and the punchline is that you lose the election!!民眾提出的要求也還是人口增加社會的思維,真的很難吧,就算做出正確的政治判斷,也不會被接受……最後的結局就是選舉落選!!
Answered已回答回答済みWritten response書面回覆書面回答
When a society is no longer growing, continuing to add requests makes prioritisation harder. Even an unpopular choice needs a record of who decided, on what grounds, what was protected and how it can be reviewed later. A receipt lets a decision remain legible beyond the next election, whether or not it wins immediate approval.
自治体のローカルルールの課題について検討していますが、自治体側から警戒されることが多い印象です。自治体側と連携したりヒアリングするうえで気を付けた方がよいポイントがあれば教えていただけますでしょうか。I am looking into the problems with municipalities’ local rules, but my impression is that municipalities are often wary. If there are points worth watching out for when working with municipalities or interviewing them, could you tell us?我正在研究地方政府地方性規則的課題,但印象中常常會受到地方政府那一端的戒備。如果在與地方政府合作或進行訪談時有什麼該注意的重點,可以請您分享嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Wariness can reflect a municipality’s duty to protect residents and staff, not a closed mind. Begin smaller than the trust being requested: agree on one purpose, the data involved and the result to return, with an easy opt-out. Before asking for trust, offer a procedure that can be inspected and reversed.
先ほどの質問にadd onして、例えば追加情報や背景をユーザーに提供できるようなAIアシスタントを投入することに対してはどう思われますか?Adding on to the earlier question, what do you think about introducing an AI assistant that could, for example, provide users with additional information and background?想接續剛才的問題再 add on 一下,例如導入能夠提供使用者額外資訊與背景的 AI 助理,您對此有什麼看法?
Answered已回答回答済みWritten response書面回覆書面回答
An AI assistant can help by supplying context and related public information. It should not nudge people towards a preferred question or manufacture a conclusion before the microphone. Agree first on what it may add, let people set the question, and keep its sources and suggestions inspectable. That links directly to the 13:30 municipal session in the same hall, where Kurimoto-san showed how every voice can receive a return.
AI 助理可以補充資料背景與相關公開資訊,作為理解脈絡的協助。但不能誘導提問,也不能在開口前替人製造結論。先約定它能補充什麼,由人決定提問,並讓使用的來源與建議事後都可查驗。這也直接連到同一個第 2 會場、13:30 的地方政府場次,栗本さん在那裡示範了如何讓每個聲音都收到回應。
Your test reached us safely, and sending a question, receiving an answer, and verifying it here demonstrates one complete loop. Thank you for checking the channel with us, so the same simple, visible receipt can remain available when the questions carry more weight.
各個人が(AIインタビュー入れるなどして)「やりたいこと」を明確化することは市民参加ではないですよね? 「自分が」ではなくて、この社会が、私たちのために、どのような施策やルールがあった方がいいかを考えねばならないはずで、そもそもの議論の前提からおかしくありませんか?Each individual clarifying “what they want to do” (by bringing in AI interviews and so on) is not civic participation, is it? Rather than “what I want”, we ought to be thinking about what policies and rules this society should have for us, so isn’t the very premise of the discussion off?每個人(透過導入 AI 訪談之類的方式)把「想做的事」明確化,這並不算公民參與吧?重點不該是「我自己」,而是必須思考這個社會為了我們,應該要有什麼樣的政策與規則,所以整個討論的前提本身不是就有問題嗎?
Answered已回答回答済みWritten response書面回覆書面回答
You are right that clarifying a person’s wishes through an AI interview is not, by itself, civic participation. The shared question must be set by people, and the AI should help someone articulate experience rather than manufacture an opinion. Its output then belongs in human-to-human deliberation, where rules for our common life can be considered.
你說得對。若只是透過 AI 訪談釐清個人的「想做的事」,本身還不能算市民參與。共同的問題必須由人來設定,AI 只協助把經驗說清楚,而不是製造意見,最後還要回到人與人之間的熟議,思考我們共同生活所需的規則。
AIインタビューやしゃべれるマニフェストのような対話による意見の引き出しは、ともすると意見誘導や意見の刷り込みという批判もありうるように思いますが、その信頼性をどうすれば担保できそうですかね?Drawing out opinions through dialogue, as with AI interviews or a talking manifesto, could easily invite criticism of steering opinions or implanting them; how do you think its trustworthiness could be guaranteed?像 AI 訪談或會說話的政策綱領這種透過對話來引出意見的做法,我覺得很可能會招來「誘導意見」或「灌輸意見」的批評,請問要怎麼做才能確保它的可信度呢?
Answered已回答回答済みWritten response書面回覆書面回答
That criticism is valid. Nakayama-san himself acknowledged today that improving the quality of a voice can also change the speaker’s thoughts. Trust therefore starts before the microphone: explain the use and obtain agreement, let humans set the questions, make prompts and logs inspectable, and offer an opt-out. The AI should draw out experience, not plant an opinion.
AIインタビューを試しに以前利用しましたが、自身の意見がますます拡張するかのような質問ばかりされ、恐ろしく思いました。とはいえ、AIに反対意見や批判を述べさせることも技術的には可能かもしれないが、離脱を招くでしょう。無用どころか、他者を呼び込まないリスクをAIインタビューは増大させませんか?I tried out an AI interview once before, and I found it frightening that it asked nothing but questions that seemed to expand my own opinion further and further. That said, while it might be technically possible to have the AI state counterarguments or criticism, that would cause people to drop out. Far from being merely useless, doesn’t the AI interview increase the risk of not drawing others in?我以前曾經試用過 AI 訪談,結果被問的全都是彷彿會讓自己的意見不斷擴張的問題,讓我覺得很可怕。話雖如此,就算技術上或許可以讓 AI 說出反對意見或批評,那也會導致中途退出吧。AI 訪談豈止是沒有用,難道不會反而增加無法把他人拉進來的風險嗎?
Answered已回答回答済みWritten response書面回覆書面回答
That fear is an important warning from lived use. An AI that keeps agreeing may amplify a view, while forcing a counterargument can make people leave. The remedy is not to stage artificial balance. Before the microphone, explain the purpose and obtain agreement; publish the questions and prompts; allow skipping and withdrawal. Trust grows from genuine room to choose, including the choice not to continue.
興味のない人のふわっとした意見の中から、もしかしたら誰にも注目されないしかし重要な指摘や少数の立場を代弁するような質的な声を拾うことが、AI等のテクノロジーの役割なのではないのでしょうか。自己理解を期待し、人様の意見を生煮えばかりだと断定することはやめた方がいいのでは?Isn’t it the role of technology such as AI to pick up, from the vague opinions of people who are not interested, the qualitative voices that might go unnoticed by anyone and yet make an important point or speak for a minority position? Rather than expecting self-understanding, shouldn’t we stop declaring that other people’s opinions are nothing but half-cooked?從沒有興趣的人那些模模糊糊的意見當中,把可能沒有人會注意到、卻很重要的指摘,或是能為少數立場發聲的質性聲音撈出來,這不正是 AI 等科技該扮演的角色嗎?一邊期待自我理解,一邊斷定別人的意見全都只是半生不熟,這種做法是不是停止比較好呢?
Answered已回答回答済みWritten response書面回覆書面回答
Yes. Surfacing an overlooked experience or an important signal from a hesitant, loosely formed voice is precisely where broad listening earns its place. We should not label such contributions half-cooked and discard them. The AI can surface candidates, while people check them against the original words. It must never pretend to represent everyone; it should help us see who is missing.
熟慮と真逆の「倍速」で、事前にわかっていた「この辺りだろう」に落ち着くことを自己理解と呼ぶことのどこに正統性があるのでしょう?Where is the legitimacy in calling it self-understanding when, at “double speed” that is the exact opposite of careful reflection, one settles on the “roughly this” that was already known in advance?用與深思熟慮完全相反的「倍速」,最後落回到事前就已經知道的「大概就是這一帶吧」,把這種情況稱為自我理解,正當性究竟在哪裡呢?
Answered已回答回答済みWritten response書面回覆書面回答
Your doubt is justified. If acceleration merely returns someone to a conclusion already expected, calling it self-understanding has no legitimacy. An AI summary can be a provisional mirror, but the person must choose questions, revise words, and withdraw consent. Legitimacy for a public measure comes from human deliberation and an inspectable route, not from speed.
AIに自身の思考すら無意識的に託してしまうリスクはあるってことですかねSo there is a risk that we unconsciously entrust even our own thinking to AI, is that what it means?所以意思是說,確實存在著在無意識中連自己的思考都託付給 AI 的這種風險,是嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, that risk is real. As we answer fluent questions, it can become unclear whether a thought came from us or was gently supplied by the system. The safeguards are prior agreement about use, visible questions and prompts, and the freedom to revise or leave. AI should not think on our behalf. It should be a mirror that returns experience to us.
層化抽出されたこの熟議の参加者には日当はどの程度出たのでしょうか?Roughly what level of daily allowance was paid to the participants in this deliberation, who were selected by stratified sampling?這場審議的參與者是以分層抽樣選出的,請問發給他們的每日津貼大約是多少呢?
The precise daily allowance belongs to the organisers of this deliberation, so I will not guess. As a principle, paying for participants’ time is part of the duty of collecting people’s voices with care. The allowance and other costs should be published alongside what changed. That makes the process’s receipt visible, rather than asking participation to run on goodwill alone.
市民熟議は大変面白い試みであると考えます。これがもっと社会に浸透すればと思っておりますが、それにあたって、全体的な費用(予算)はどれぐらい掛かったのでしょうか。差支えない範囲で、概算という形で教えていただけないでしょうか。I consider citizen deliberation a very interesting experiment, and I hope it will spread further through society. On that note, roughly how much did the whole exercise cost (as a budget)? Could you share a rough estimate, to the extent you are free to do so?我認為公民審議是非常有意思的嘗試,也希望它能更加深入社會。想藉此請教,整體費用(預算)大約花了多少呢?在方便透露的範圍內,能否以概算的形式告訴我們?
The organisers should confirm the overall estimate, rather than my inventing a figure. The principle is clear: allowances, venue, briefing materials, facilitation, records, and analysis all belong on the same receipt as the results. Publishing what was spent and what changed makes deliberation a public good that others can understand and adapt.
“熟議一般”というよりも、討論型世論調査など“熟議制度としての熟議”のお話と理解しました。台湾型だけでなく、欧州の気候市民会議や憲法改正などで行われている先行する熟議制度の中で、日本に今取り入れたいと考えている仕組みはありますか?あるいは、ここは日本とは合わなそうだ、という点があれば教えてお伺いしたい。I took this to be about “deliberation as a deliberative institution”, such as Deliberative Polls, rather than “deliberation in general”. Among the deliberative institutions that have gone before us (not only the Taiwan model, but those used in Europe’s climate citizens’ assemblies, constitutional amendment and so on), is there a mechanism you would like Japan to adopt now? Or, if there are points where you feel these would not suit Japan, I would like to hear them.我的理解是,這裡談的與其說是「一般的審議」,不如說是審議式民調等「作為制度的審議」。在不限於台灣模式、還包括歐洲氣候公民會議與修憲等已有先例的審議制度當中,有沒有您現在就想引進日本的機制?反過來,若有覺得「這一點恐怕與日本不合」之處,也想請教。
Answered已回答回答済みWritten response書面回覆書面回答
Since you asked for my own pick: I would start with a permanent, sortition-based Citizens’ Council, not a one-off assembly, whose members track Parliament’s response and implementation, returning receipts to the public. Ostbelgien has run this as a law-mandated permanent institution since 2019. In my personal view, a process that rushes towards a “beautiful consensus” or stages unanimity does not suit Japan well. That is my own opinion, not an imported correct answer.
熟議に当たって、その方が保有する基礎知識によって理解度に差が出ると思いますが、知識の底上げのような活動内容は不要でしょうか?In deliberation, I imagine understanding will vary with the basic knowledge each participant holds. Is there no need for activities that raise the floor of knowledge?在審議時,理解程度想必會因每個人具備的基礎知識而有差異;請問不需要「拉高知識底線」之類的活動內容嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, a knowledge floor matters if deliberation is to be fair. The briefing should give everyone the same starting materials, prepared with care and presenting genuinely neutral options, as today’s panel described. People should be able to ask questions before deciding, and the materials themselves should be public and open to scrutiny. Equal access to understanding comes before equal speaking time.
熟議においては、声が大きい人、ロジカルな考え方ができる人がいるかいないかで結果が変わってき得るように思ったのですが、いかがでしょうか?In deliberation, it seemed to me that the outcome could shift depending on whether the room happens to include loud voices or people adept at logical thinking. What is your view?在審議中,我覺得結果可能會因在場是否有嗓門大的人、或擅長邏輯思考的人而改變,您怎麼看呢?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, louder voices and quicker logic can tilt an outcome. Facilitation, small groups, speaking order, and equal-time tools can widen the doorway. Yet cl-07 is right that equal seconds alone are not justice. Fairness must also include the safety to speak, the time to reconsider, and the health of the relationships in the room. Mechanics should serve that wider fairness.
ネット上の偽誤情報対策は総務省が主管だと思いますが、この活動に対する総務省のコメント等はありましたか?I believe countermeasures against disinformation and misinformation online are under the charge of the Ministry of Internal Affairs and Communications. Has the ministry offered any comment or the like on this activity?網路上不實與錯誤訊息的對策,我想主管機關應是總務省;請問總務省對這項活動有過評論之類的回應嗎?
Whether the Ministry of Internal Affairs and Communications has offered a comment is for the government and its responsible office to answer. I will not speak in its place. The structural principle is clearer: an exercise needs independence in how questions are set, people are included, records are kept, and answers return. Its legitimacy should not depend on a ministry’s blessing.
時間の平等さだけでは、公平かもしれないが公正ではないこともあるのでは。逆に、こういったテクノロジーによる話し合いへの機械的な介入に慎重さも必要ではないでしょうか。Equal time alone may be fair, but is there not a sense in which it is still not just? Conversely, don’t we also need caution about this kind of mechanical intervention by technology in people’s discussions?僅靠時間上的平等,或許算公平,卻也可能並不公正,不是嗎?反過來說,對這種由科技對談話所做的機械式介入,是否也需要保持謹慎呢?
Answered已回答回答済みWritten response書面回覆書面回答
Your caution is right. Equal speaking time does not create equal confidence, room to rethink, or permission to ask again. We can use tools to address the imbalance raised in cl-05, but the machine must not decide what fairness means. Human facilitation and participant consent stay central. Technology is a vehicle for healthier relationships, not a mechanical substitute for them.
本セッションでは、Slido は利用しませんSlido will not be used in this session.本場次將不使用 Slido。
Answered已回答回答済みWritten response書面回覆書面回答
Choosing not to use Slido can be right for a room, whether because of time or the sensitivity of the topic. It does not make the questions disappear. A loop closes when a voice receives an address, and this page is that address: a small, public receipt that the question arrived and was answered.
LLMポイズニング 為政者にとっては、かなり強力な武器になりますね。この話はもっと国民や学校の子どもたちに伝えないといけないと思いますが、どうでしょうか?LLM poisoning: for those in power, it becomes quite a powerful weapon, doesn't it. I think this needs to be told far more widely to the public and to children in schools. What do you think?LLM 毒化,對執政者而言會成為相當強大的武器呢。我認為這件事必須更多地傳達給民眾和學校裡的孩子們,您覺得呢?
Answered已回答回答済みWritten response書面回覆書面回答
Yes. Preparing for LLM poisoning means learning to trace where information comes from and how it travels, not memorising approved answers. Schools are the right place for provenance literacy and prebunking: showing how contamination enters an information environment before it spreads. Children should be treated as participants in keeping the civic well clean, not merely as people to protect.
中国のような権威主義国はどのように介入すべきナラティブを選定しているのかが気になったI was curious how authoritarian states like China go about selecting which narratives they should intervene in.我很好奇像中國這樣的威權主義國家,是如何選定應該介入的敘事。
Answered已回答回答済みWritten response書面回覆書面回答
The research pattern is broader than blaming any one country. Information operations tend to amplify some narratives while leaving gaps in a language space, including Japanese-language spaces. Our answer is plural sources, visible provenance and open evaluation: people should be able to inspect not only the story, but also who selected it and how its reach was assessed. The selection mechanism itself can be traced through attribution research with open methods, so judge reports by verifiability rather than by conclusions drawn from country names.
紙も併用されたとのことですが、アプリにはどのように反映されたのでしょうか?I understand paper was used in combination as well. How was that reflected in the app?聽說也併用了紙本,請問那些內容是如何反映到 App 上的呢?
The precise implementation belongs to Ariga-san and the makers, so I will not invent a product roadmap. The principle is hybrid respect: paper is where people can think, revise and compare together; an app can then help organise and share, with consent. Keeping both traces makes it possible to check how each voice travelled.
社会に関する対話の楽しさの設計という点は、非常に重要だと思います。政治の話は対立が起きがちですが、対面そのものには対立を避ける効果があるように思えます。一方で、匿名だからこそ吐き出せる本音もあると思うのですが、匿名の完全オンライン対話でうまく実現できる設計はありますでしょうか(完全な匿名性はなくて良いと思いますが)I think designing enjoyment into dialogue about society is extremely important. Political conversation tends towards conflict, yet meeting face to face seems in itself to help avert it. On the other hand, I think there are honest feelings people can voice precisely because they are anonymous. Is there a design that can achieve this well in fully online, anonymous dialogue? (I don’t think complete anonymity is necessary.)「為社會議題的對話設計樂趣」這一點,我認為非常重要。政治話題容易引發對立,而面對面本身似乎就有避免對立的效果。另一方面,我想也有正因匿名才能一吐為快的真心話;請問有沒有能在完全線上的匿名對話中妥善實現這點的設計?(我認為不需要完全的匿名性。)
Answered已回答回答済みWritten response書面回覆書面回答
Anonymity and warmth do not have to be opposites. We can use pseudonyms, keep groups small and agree the rules before anyone speaks. The organisers may retain limited records for safety, while publication remains a choice for the person speaking. That leaves room for honest disclosure without giving up the human feeling that a relationship is forming.
タウンミーティングは支持者向けのワークショップになるのかなと思います。支持者以外(無党派層含む)にはどんな要素を大事にしてリーチしていくことが必要だと思われているでしょうか?またどんなアウトプットをチームみらいはしていくんでしょうか?I suspect the town meetings will turn into workshops for existing supporters. In reaching people beyond supporters (including unaffiliated voters), what elements do you consider it important to focus on? And what outputs will Team Mirai go on to produce?我猜這類市民座談會(town meeting)可能會變成以支持者為對象的工作坊。要觸及支持者以外的人(包括無黨派族群),您認為需要重視哪些要素?另外,Team Mirai 今後會端出什麼樣的產出呢?
Answered已回答回答済みWritten response書面回覆書面回答
A town meeting should not quietly become a proxy for the people who already support it. Sortition, practical access such as childcare and interpretation, and venues people already trust can widen the invitation. I would look for a public receipt showing shared values, unresolved questions and how the material was handed onwards. The team’s concrete outputs belong to them; the transparency principle is ours to carry.
不能只靠參與者的熱情,就把一個只傳達支持者聲音的座談會當成人民的代表。可用隨機抽樣,補上托育、口譯等參與條件,也走進圖書館、學校和社區聚會等本來就受信任的地方。成果不只統計贊成反對,更要用公開收據呈現共同價值、未解決的問題,以及如何交給政策。具體產出由 team みらい 說明,但透明原則由我們守住。
The app’s current release status and roadmap belong to its makers, so I will not guess. Today, Anno-san offered an open-source AI interview and open-data API approach, explicitly including other parties. That spirit answers this request: public tools should be inspectable, reusable and improvable, including by municipalities. This page is the receipt and the route back to the makers.
關於 App 目前的發布路線,應由開發者確認,我不替他們猜測。今天安野さん提出開放原始碼的 AI 訪談,以及納入其他參與者的開放資料 API 做法,正回應了這項要求:公共工具應可檢視、重複使用,也應讓包含地方政府在內的任何人持續改進。這個頁面就是送回開發者手中的收據與路徑。
座談会アプリの公開状況や予定は、製品をつくる登壇者の方に確認するのが正確です。今日、安野さんが他党の方も含めて AI インタビューとオープンデータ API の公開を呼びかけた姿勢は、この問いにも響きます。自治体が検証し改良できる OSS の可能性を、このページを宛先として作り手へ届けます。
mw-05Anonymous▲1
何か課題、今後に向けた改善点はないですか?Are there any challenges, or points you would improve going forward?有沒有什麼課題,或今後想改進的地方呢?
Answered已回答回答済みWritten response書面回覆書面回答
The central challenge is not opening a dialogue once. It is lowering the burden of participation while showing where each voice goes. We should record who could not join, how paper and speech were organised and what did not reach a decision. Improvement therefore means clearer purpose before participation, a receipt afterwards and the ability to correct or leave along the way.
AIインタビューや座談会アプリもオープンソース化していただきたいです自治体でも活用したいですI would like the AI interview and the roundtable app to be open-sourced too. We would like to use them in municipalities as well.希望 AI 訪談與座談會 App 也能開源。我們在自治體也很想使用。
That request is concrete: municipal reuse is one of the best tests of an open-source civic tool. The app’s release scope, licence and support plan belong to the makers, so I will not invent them. Alongside Anno-san’s offer to include other parties, this page carries the municipal use case back to the people who can answer it.
“I would like to try it” is already useful civic feedback. The makers should answer about release plans, but Anno-san’s open-source offer points towards a healthy cycle: more people try the tool, municipalities and communities adapt it and improvements return to everyone. This page is the receipt that the wish reached its makers.
デジタル民主主義とは関係ありませんが、それまでの業務などをDXで代替していく上で、「DXなんだから紙は使うな」というような運用がされることもあると聞きます。デジタルデバイスや技術が現実の制約をエクステントするものとして使われるために、それらの導入にはどのような心持ちが必要だと思われますか。This is unrelated to digital democracy, but I hear that when existing work is being replaced through DX, operations are sometimes run on the line of “this is DX, so no paper allowed”. For digital devices and technologies to be used as things that extend the constraints of the real world, what mindset do you think is needed when introducing them?這雖然與數位民主無關,但聽說在以 DX(數位轉型)取代既有業務的過程中,有時會出現「都已經是 DX 了,就別再用紙」這樣的運作方式。為了讓數位裝置與技術被當作延展現實限制的工具來使用,您認為在導入時需要什麼樣的心態呢?
Answered已回答回答済みWritten response書面回覆書面回答
DX should extend human choices, not amputate familiar ones. As Osugi-san kept respect for paper and for the staff who make the final judgement, paper should remain available where it carries accessibility or trust. Digital tools can add search, sharing and traceability. The right mindset is to expand options while keeping responsibility visible.
みらいは明るいと信じられる国 でポジティブな意見例があったのかFor “a country where people can believe the future is bright”, were there examples of positive opinions?在「能夠相信未來是光明的國家」這一題中,有出現過正面意見的例子嗎?
I cannot responsibly invent or select a representative positive comment without checking the session record. The wish for a country where people can believe the future is bright is itself a meaningful policy criterion. A receipt should preserve the speaker’s words, what changed and what remains open, so hopeful views are returned to the next dialogue rather than consumed. Ariga-san’s packaging ambition can widen that cycle.
これらの声がどのように政策に反映されていくのか気になりますI am curious how these voices will go on to be reflected in policy.很好奇這些聲音今後將如何反映到政策之中。
Answered已回答回答済みWritten response書面回覆書面回答
Voices reach policy only when the path before and after a decision is visible. A receipt should show what was said, who used it, when it informed a decision and whether it was adopted, held or declined, with reasons where possible. The decision itself belongs to politics and administration. Ariga-san’s packaging ambition can make that handover easier to follow.
能動的な行動を必要とするモデルなので、どのように市井の方のモチベーションを高めるのかがネックになるかと思いますが、それに対してはいかがお思いでしょうかSince this model requires people to take the initiative, I suspect the sticking point will be how to raise the motivation of ordinary citizens. What are your thoughts on that?這是一個需要人們主動行動的模式,我想「如何提升一般民眾的參與動機」會是瓶頸所在,對此您怎麼看呢?
Answered已回答回答済みWritten response書面回覆書面回答
Motivation grows when participation has both joy and consequence. A person speaks, sees where the thought travelled, notices it shape a decision or a new question and receives a reply. That closed loop is fun in the best sense. Playful facilitation helps, but the lasting invitation is a visible receipt and a clear next step for ordinary people.
テーマをこまやかにした対話の場を設けようとするとき、そのテーマの当事者に対してどのように場の存在を知ってもらうのでしょうか?When you set out to create dialogue spaces on more finely grained themes, how do you make the people directly concerned by a theme aware that the space exists?當要針對更細緻的主題開設對話場域時,要如何讓該主題的當事人知道這個場域的存在呢?
Answered已回答回答済みWritten response書面回覆書面回答
Do not wait for people affected by a topic to discover a recruitment page. Go to places they already gather, including workplaces, schools, local groups and trusted small online spaces. Explain the purpose and participation terms before asking for a voice, then return a receipt showing what changed. Trusted entry points and visible exits belong together.
賛成、中間、反対という切り取り方で場を設定するよりも、賛否や無関心のなかに共通する価値をどう発見するか、ということなのではないか?Rather than setting up the space by carving people into for, neutral and against, isn’t the point how we discover the values shared across support, opposition and indifference?與其以贊成、中間、反對這樣的切分方式來設定場域,重點難道不在於「如何在贊成、反對與漠不關心之中發現共通的價值」嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Yes. That is the heart of deliberation. Supporter, undecided and opponent can be useful labels, but they flatten the wishes underneath. If we first find values that survive across positions, such as safety, dignity, participation and continuity, we can make the next judgement together without erasing disagreement. Discovering that common ground is how today’s voices return to policy.
品川区の取組について。標準的な役所マインド(元行政関係者です)だとわざわざ行政評価などで目立ちたくないと思いますが、外部評価に20事業も手が上がるというのはどういう設計上の考慮があったのでしょうか?About Shinagawa Ward’s initiative. With the standard municipal-office mindset (I used to work in public administration), one would rather not stand out in exercises like administrative evaluation. What design considerations lay behind as many as twenty projects putting themselves forward for external evaluation?想請教品川區的做法。按照公務機關的標準心態(我曾是行政相關人員),照理不會想特地在行政評估這類場合出風頭;究竟是有什麼制度設計上的考量,才會有多達 20 項事業主動舉手接受外部評估呢?
The details of why twenty Shinagawa projects volunteered belong to the speakers from Shinagawa. As a general principle, when evaluation is designed as help rather than punishment, and being seen feels safe, visibility can become an asset rather than a burden.
行政の都合で「オープンな」声の受付が行われているように見えます。これがデジタル民主主義なのでしょうか?It looks as though an “open” intake of voices is being run at the administration’s convenience. Is this digital democracy?看起來,「開放的」意見受理像是配合行政方的方便在進行。這樣就是數位民主了嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Your concern is right. If an administration collects voices for its own convenience without agreement about their use, that is extraction, not yet digital democracy. The remedy is an engagement contract and participation upstream, before the microphone. Then the answer must return: what was heard, what changed and who decided. Openness becomes democratic when the loop closes.
集まった意見については、言い方が難しいですが「意識の高い」人のものである可能性が高いのではないかと思いますところ、「意識の低い」人の意見はどのように収集されている(あるいはされたい)でしょうか?As for the opinions collected (though it is difficult to phrase), I suspect they are very likely to come from the “highly engaged”. How are the opinions of the “less engaged” being collected (or how would you wish them to be)?關於蒐集到的意見,雖然不知該怎麼說才好,但我想很有可能多半來自「意識較高」的人;請問「意識較低」的人的意見,是如何蒐集的(或者希望如何蒐集)呢?
Answered已回答回答済みWritten response書面回覆書面回答
Behind “less engaged” is a fair warning that our entrances are uneven. We can use sortition, meet people where they already are, and keep paper and face-to-face options alongside digital channels. As Nakayama-san said today, the honest promise is breadth, not guaranteed representation. We should keep checking whose voice is still missing, rather than declaring the sample complete.
デジタル技術の登場で変わったこと、変わっていないことはどのようなところでしょうか。世田谷の学生の方の取り組みは、しばらく前から鋭意なされている若者の声反映の取り組みのように思えたもので、御意見うかがえればと思います。What has changed with the arrival of digital technology, and what has not? The Setagaya students’ initiative struck me as the kind of work to reflect young people’s voices that has been pursued in earnest for some time now, so I would be glad to hear your views.數位技術出現之後,哪些改變了、哪些沒有改變呢?世田谷學生們的行動,在我看來像是一段時間以來便持續認真推動的青年意見反映工作,希望能聽聽您的看法。
Answered已回答回答済みWritten response書面回覆書面回答
Digital tools make it easier to gather, organise and show the path from a contribution to a response. They do not create the trust that lets young people speak, nor the responsibility to listen. Setagaya’s work is a useful bridge: add digital reach to practices already rooted in relationships, while keeping face-to-face and paper channels available.
住民と自治体だけでなく、交通機関の意見は参考にしたのでしょうか?例えば若者向けの助成制度により混雑が発生する可能性やこの対策としてのダイヤ見直し等彼らの意見も聞いた方がいいように思います。Beyond residents and the municipality, did you also consult the views of the transport operators? For example, the possibility that a subsidy scheme for young people causes crowding, and timetable revisions as a countermeasure; it seems to me it would be better to hear their views on such points too.除了居民與自治體之外,是否也參考了交通業者的意見呢?例如青年補助制度可能造成的擁擠,以及作為對策的班次時刻調整等,我覺得也應該聽聽他們的意見。
I cannot confirm from here whether transport operators were included, so that detail belongs to the Shinagawa speakers. The principle is clear: anyone affected by a subsidy, including the people who keep services running and nearby users, should enter upstream. Their evidence should shape mitigation, and the final decision should come back with reasons.
行政評価を、なぜ始められたのでしょうか? 予算内におさまった、計画したものを完成したといった事実があっても、区民から不満などがあったのでしょうか?Why did you begin administrative evaluation in the first place? Even with the facts that spending stayed within budget and planned projects were completed, was there nonetheless dissatisfaction from ward residents?當初為什麼會開始推動行政評估呢?即便有「支出控制在預算內」「完成了既定計畫」這樣的事實,是否仍存在來自區民的不滿呢?
The history of Shinagawa’s evaluation and whether residents expressed dissatisfaction belongs to its speakers. The principle is to ask more than whether a project stayed within budget or finished on schedule: whom did it reach, and whom did it miss? Dissatisfaction is not a verdict. It is information that can guide the next repair.
しなトークで集めた意見について、①投稿者の属性や流入経路の偏りをどう把握するのか、②賛否や少数意見をどの基準で整理・統合するのか、③非公開の行政評価委員会での判断過程と、反映・不採用の理由をどこまで公表する予定か、教えてください。Regarding the opinions gathered through ShinaTalk, please tell us: ① how you will grasp biases in contributors’ attributes and in the routes by which they arrive; ② by what criteria you will organise and consolidate support, opposition and minority views; and ③ how far you plan to disclose the decision process of the closed administrative evaluation committee, and the reasons for adopting or not adopting ideas.關於透過「しなトーク」(ShinaTalk)蒐集的意見,想請教:①如何掌握投稿者屬性與流入管道的偏差;②以什麼標準整理、統整贊成、反對與少數意見;③非公開的行政評估委員會之判斷過程,以及採納與不採納的理由,預計公開到什麼程度。
The speakers must confirm Shinagawa’s exact publication plan. My principle is to return all three parts: the attributes and routes that may bias the inflow; the criteria used to combine or separate supportive, opposing and minority views; and the reasoning of the closed committee, including why an idea was adopted or not. A receipt says who spoke, what changed and who answered.
そもそも意見を集めることを「答えを出す」に結びつけねばならない必然性があるのでしょうか?特定の政策課題というフォーカシングを行政がおぜん立てする限り、それ自体の抱えるバイアスから自由にはなれません。むしろもっと緩く場を設定して談義をさせ、そこからさまざまな観点を抽出するという発想が必要ではないかと思います。Is there really any necessity, in the first place, to tie collecting opinions to “producing an answer”? So long as the administration stages the focus on a particular policy issue, the exercise cannot be free of the bias that framing itself carries. Rather, I think what is needed is the idea of setting the scene more loosely, letting people talk at large, and drawing out diverse perspectives from that.說到底,把蒐集意見與「得出答案」綁在一起,真的有其必然性嗎?只要「聚焦於特定政策課題」這件事是由行政機關來鋪陳,就無法擺脫其本身所帶的偏差。我反而認為,需要的是更寬鬆地設定場合、讓大家自在漫談,再從中萃取各種觀點的想法。
Answered已回答回答済みWritten response書面回覆書面回答
No. Producing a single answer is not the only legitimate outcome. If the administration fixes the policy question too early, its framing bias remains. Loose spaces for conversation can surface perspectives we did not know to ask for. The loop can close when we say, honestly, “Having listened, our question changed.”
品川区さんでは、色々な意見を踏まえ、事務事業整理の最終的な意思決定はどのようにされているのでしょうか?How does Shinagawa Ward, having taken the various opinions on board, make the final decisions on sorting out its administrative programmes and projects?請問品川區在參酌各方意見之後,事務事業整理的最終決策是如何做成的呢?
The exact decision flow and authority belong to Shinagawa’s speakers, and I do not want to invent them. The principle is that the administration should show how contributions were organised, who made the final choice and why. People need more than an outcome. They need a public decision trail they can inspect and question.
行政評価というと、旧民主党政権の事業仕分けを思い出してしまいます。あれはコスト削減アピールの政治ショーの要素が強い取り組みでした。行政評価のフローの中に住民の声をいれていくというアプローチは素晴らしい取り組みだと思います。ぜひ、政治家のおもちゃにならないような取り組みにしていってください!Administrative evaluation reminds me of the budget screening (jigyō shiwake) under the former Democratic Party of Japan government. That was an exercise with a strong element of political theatre, staged to advertise cost-cutting. I think bringing residents’ voices into the flow of administrative evaluation is a wonderful approach. Please make sure it never becomes a plaything for politicians!說到行政評估,我不禁想起舊民主黨政權的「事業仕分け」(事業甄別)。那是一場政治秀色彩濃厚、主打宣傳刪減成本的活動。而把居民聲音納入行政評估流程的做法,我認為非常好。請務必別讓它淪為政治人物的玩具!
Answered已回答回答済みWritten response書面回覆書面回答
Your caution about the memory of the administrative review is understandable. Bringing residents’ voices into evaluation must be about winning trust, not surveillance or attack. I would keep Ono-san’s line from today at the centre: “win trust rather than monitor and attack.” Publish the criteria, reasons and room for repair, so evaluation cannot become political theatre.
藤原さんの「仲介」マインドが目覚めたきっかけをお伺いしたいです。媒介者が必要というのはとても良い気づきだと思います。I would like to ask what first awakened Fujiwara-san’s “mediation” mindset. The realisation that go-betweens are needed strikes me as a very good insight.想請教喚醒藤原先生「仲介」心態的契機是什麼。「需要媒介者」這一點,我認為是非常好的體悟。
Fujiwara-san’s origin story belongs to Fujiwara-san, and I would rather hear it in his own words. The principle is warm and practical: a mediator is not someone who dilutes a view, but a bridge that lets different positions meet again. By preparing the room, loosening misunderstandings and carrying the decision back, mediators build civic muscle.
①少数者の②人権擁護、というタイトルだと、当然のことながら「尊重」したいという感情を呼び起こし、「尊重」すべきだという空気感を生みます。プロセスとしての熟議をスキップしてしまう危惧はありませんか。声高に叫ばれるISSUEが「人権」と当然に認めて良いのか、「声の大きな少数者の主張」が寡多数派の道徳観や常識を覆すのはありかWith a title like “① minorities’ ② protection of human rights”, it naturally stirs the feeling of wanting to “respect” them and creates an atmosphere that one “ought to” respect them. Is there no risk that deliberation as a process simply gets skipped? Can an ISSUE that is loudly proclaimed be accepted as a “human right” as a matter of course, and is it acceptable for “the claims of a vocal minority” to overturn the moral sense and common sense of the silent majority?若標題是「①少數者的②人權保障」,自然會喚起想「尊重」的情感,並形成「應該尊重」的氛圍。難道不擔心因此跳過了作為程序的審議嗎?被高聲疾呼的 ISSUE,可以理所當然地被承認為「人權」嗎?「聲量大的少數者的主張」推翻沉默多數的道德觀與常識,這樣可以嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Human rights cannot depend on winning a deliberative popularity contest. Courts exist precisely so that dignity has a remedy when a majority is silent or mistaken. Deliberation still matters: it tests facts, reasons and consequences in public. Rights protection and deliberation are not rival camps. They are complementary procedures, each keeping the other honest.
デジタル民主主義が進めば、最終的にはそこから立法までにもつながりうると思ういます。その際の展望と限界について教えていただきたいですIf digital democracy advances, I think it could ultimately lead all the way to legislation. I would like to hear about the prospects and the limits of that.如果數位民主持續推進,我想最終甚至可能一路連結到立法。想請教屆時的展望與侷限。
Answered已回答回答済みWritten response書面回覆書面回答
Digital participation can inform legislation, but a tally of voices does not become law by itself. Legislators and administrators must explain what they took up, what they did not and how any change entered the text. The limits are real: participation can be uneven, and deliberation cannot be skipped. A public decision trail makes both visible.
公共訴訟のお話を伺っていると団体訴訟が活用できた方が、市民にとってメリットが大きいと考えますが、先の国会での個人情報保護法改正を見ても団体訴訟の普及は課題が大きい模様です。公共訴訟における団体訴訟の活用の可能性はどう考えておられますか?Listening to the discussion of public litigation, I believe citizens would benefit more if group litigation could be put to use. Yet judging from the amendment of the Personal Information Protection Act in the recent Diet session, spreading group litigation appears to face substantial hurdles. How do you view the potential for using group litigation in public litigation?聽了關於公共訴訟的討論,我認為若能善用團體訴訟,對民眾的好處更大;但從上個國會會期《個人資料保護法》的修法來看,團體訴訟的普及似乎面臨很大的課題。請問您如何看待在公共訴訟中運用團體訴訟的可能性?
The legal availability and legislative prospects of group actions belong to CALL4’s legal practitioners, so I will not invent an answer. The principle is to make the doorway wider: people should not have to carry a shared issue alone, common facts should become visible and a judgment should return as a reason to improve the system. Collective action is a means, not the goal.
訴訟はどの時点から公共訴訟になるのか。ある人やグループによる訴訟が、どのような状態になると、公共訴訟として扱われるのか。At what point does litigation become public litigation? When a lawsuit is brought by a person or a group, under what conditions does it come to be treated as public litigation?訴訟從哪個時點起算是公共訴訟?由個人或團體提起的訴訟,要達到什麼樣的狀態,才會被視為公共訴訟呢?
Answered已回答回答済みWritten response書面回覆書面回答
Public litigation is not defined only by the number of claimants or the attention a case receives. Public significance appears when an individual rights dispute also concerns people living under the same rule, and when the court’s reasoning can test a wider practice. The precise legal classification belongs to CALL4’s practitioners, not to an improvised definition here.
訴訟を起こす、というのは、時間とエネルギーが必要なアクションのようなイメージがあるが、そのハードルを下げるにはどうすればいいか。むしろハードルが下がることで起きる弊害があると思うか。Bringing a lawsuit has the image of an action that demands time and energy. How can that barrier be lowered? And do you think, on the contrary, that harms would arise precisely because the barrier is lowered?提起訴訟給人一種需要投入大量時間與精力的印象;要怎麼做才能降低這道門檻?反過來說,您認為門檻降低後反而會產生弊端嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Lowering the barrier means plain information about where to seek advice, likely costs, time and options, plus trusted links between people and lawyers. A wider door also needs safeguards: clear criteria, support for respondents and a record of why a case proceeds. Access and care must grow together, rather than trading one for the other.
CALL4の目的に賛同し入会を希望する場合、会員になれますか? 定款に記載のある、入会金と年会費はおいくらですか?If someone endorses CALL4’s aims and wishes to join, can they become a member? And how much are the admission fee and annual fee set out in the articles of association?如果認同 CALL4 的宗旨、希望加入,可以成為會員嗎?章程中記載的入會費與年費各是多少呢?
Please check CALL4’s current charter and membership desk for eligibility, joining fees and annual dues. I should not guess those figures here. The principle is warm and simple: a public-litigation platform should be open civic infrastructure, with ways for people to support it that are accessible, transparent and easy to inspect.
政党が憲法上の制度として明文化(改正)される前に、政党の存在意義が、なくなってしまうかもしれませんね。さらに、議員のありかたも変容して議員定数も、ぐっと削減できると思うのですがいかがでしょうBefore political parties are ever spelt out in the Constitution as an institution (by amendment), their very reason for being may well disappear, mightn’t it? What is more, I think the nature of being a legislator will also change, and the number of seats could be cut back sharply. What do you think?在政黨作為憲法上的制度被明文化(修憲)之前,政黨的存在意義說不定就會先消失了呢。再者,我認為議員的樣貌也會隨之轉變,議員席次也能大幅削減,您認為如何呢?
Answered已回答回答済みWritten response書面回覆書面回答
I will not take a position on whether parties should have a constitutional status or whether seat numbers should be reduced. Institutions evolve through the people’s choices and public argument. Democratic tools should serve whichever shape people choose, while keeping the essential receipt visible: who decided, on what grounds and with what result.
熟議、熟議民主主義という言い方が取り沙汰されがちだが、裁判を通じた公共訴訟というのは、闘技民主主義的に感じた。ある種、1man1voteとは別の市民社会になる希望を感じたが、これを進めていくにあたっての今後のハードルや壁は具体的に何か伺ってみたい。“Deliberation” and “deliberative democracy” are the phrases that tend to get talked up, but public litigation through the courts felt to me like agonistic democracy. In a way I sensed the hope of a civil society different from 1man1vote. I would like to ask concretely what hurdles and walls lie ahead in taking this forward.「審議」「審議式民主」這類說法常被拿出來討論,但透過法院進行的公共訴訟,讓我感覺帶有爭勝式民主(agonistic democracy)的色彩。某種意義上,我感受到一種成為有別於 1man1vote 的公民社會的希望;想具體請教,今後在推進上會遇到哪些門檻與高牆?
Answered已回答回答済みWritten response書面回覆書面回答
I like that frame. Agonism and deliberation are two rooms in one house. Litigation does not erase disagreement. It gives claims and evidence a public procedure, then leaves a judgment as a receipt of reasons. The practical walls are access and stamina: who can enter, who can afford to stay and whether the result returns to the wider public.
参加者集団のウケがいい(あるいは誰も特に強く反対しない)解を見つけることはできるけど、それを合意と呼んでよいものなのか、またそれが望ましい政策なのか、というところがまだスッキリしません。解説を加えていただけると助かります。It can find solutions that go down well with the group of participants (or that nobody opposes especially strongly), but whether that can properly be called consensus, and whether it makes for desirable policy, is still not clear in my mind. Some added explanation would be a great help.雖然能找出讓參與者群體反應不錯(或沒有人特別強烈反對)的解方,但那是否可以稱為共識、又是否是理想的政策,這一點我還沒有完全想通。若能加以解說,將會很有幫助。
Answered已回答回答済みWritten response書面回覆書面回答
A solution that feels acceptable to a room is not automatically consensus, and certainly not automatically good policy. Generative Polis can offer a floor for finding common ground, not a verdict. Legitimacy comes from who was included, what decision followed and whether a receipt shows what changed. The room’s welcome is a beginning, not the final authorisation.
讓一個房間覺得可以接受的方案,不自動等於共識,更不自動等於好政策。Generative Polis 找到的共通點,是開啟討論的起點,不是替政策下判決。只有交回一張收據,說明誰被納入、接著作了什麼決定、什麼改變,才有資格談正當性。房間的歡迎只是起點,不是最後授權。
LLMによる意見分割は、説明を聞く限り要望は吸い上げやすいかと思った一方、その背景が切り落とされがちなように思いました。本当はその背景の課題解決に別案がよい場合、本来解決すべき課題が見落とされるような気がしたが、このあたりは背景も含めうまく吸い上げてることができますか?UMAPの次元圧縮も同観点で気になりましたFrom the explanation, opinion segmentation by LLMs sounded good at gathering up requests, but I felt the background behind them tends to get cut away. If a different proposal would in fact be better for solving that underlying issue, I had the sense that the problem which really ought to be solved could be overlooked. Can this capture things well, background included? UMAP's dimensionality reduction concerned me from the same standpoint.就聽到的說明而言,我覺得用 LLM 進行意見切分似乎容易匯集需求;但另一方面,需求背後的脈絡似乎容易被切掉。如果其實另有更適合解決那些背景課題的方案,我感覺真正該解決的課題可能會被忽略。這部分能否連同背景一起妥善匯集呢?UMAP 的降維也讓我從同樣的觀點感到在意。
Answered已回答回答済みWritten response書面回覆書面回答
That concern is real. LLM summaries and UMAP projections both lose information. As Nakayama-san noted today, topic proximity can merge supportive and opposing views. We should keep each abstraction linked to the raw text, separate agreement from disagreement and prompt for background, not just topics. The compressed map is useful only when we can return to the uncompressed voice.
PolisやGenerative Polisが、コンセンサスを促すか、ポーラリゼーションを招くのか、どのようなパラメタや設定が関わるのでしょうか?What sorts of parameters and settings bear on whether Polis or Generative Polis fosters consensus or invites polarisation?Polis 或 Generative Polis 究竟會促進共識,還是招致極化,其中涉及哪些參數與設定呢?
Answered已回答回答済みWritten response書面回覆書面回答
Consensus or polarisation is shaped by choices such as the seeded statements, the wording of prompts, which similarities are shown and how often the map updates. Engagement metrics reward attention; bridging metrics ask whether people with different starting views can share a statement and explain movement. The first safeguard is to state the aim, publish the settings and keep the raw trail.
Generative Polisは、どのくらいリアルタイムな学習ができて、合意案を生成できるのでしょうか? ワークショップなどでリアルタイムな熟議に活用できるものでしょうか?How real-time can Generative Polis's learning be, and its generation of draft agreements? Is it something that could be used for real-time deliberation in workshops and the like?Generative Polis 能做到多即時的學習、進而產生共識方案呢?在工作坊等場合,它是能用於即時審議的工具嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Generative Polis can draft in real time, but trust and ratification take longer. In a workshop, I would use the live system to propose a draft, let participants inspect the linked voices and reasons, then allow time for revision and confirmation after the session. Speed is useful at the front door. It should not be mistaken for settled agreement.
Generative Polis 可以即時起草方案,但信任與批准需要更長時間。工作坊裡,可以先用即時系統提出草案,讓參與者查看連結的聲音與理由,散會後再留時間修訂與確認。速度適合當入口,不能誤當已經定案的共識。
Polis のオススメのORID が不在となっている実践と、導かれないやり方での実践での差異はありますでしょうか?Is there any difference between practice where the ORID recommended for Polis is absent, and practice carried out in a way that is not guided by it?在 Polis 所推薦的 ORID 缺席的實踐,與未受其引導的做法下的實踐之間,會有差異嗎?
Answered已回答回答済みWritten response書面回覆書面回答
An ORID sequence is not a guarantee of a good practice. It can help separate observation, reflection, interpretation and decision, making it easier to see why a proposal moved. Without that structure, the conversation may still work, but the transitions can become harder to inspect. The design question is which prompts fit the purpose, and what receipt replaces ORID when it is absent.
生の体験かあることや、音声で語られることや、信頼が準備されることで、理解された感覚か生まれ、同意が形成しやすくなる、という話かあります。そのような部分をもたせた場合の差異を測った例などありますてしょうか?There is talk that lived experience, being told in a spoken voice, and trust being prepared beforehand give rise to a sense of being understood and make agreement easier to form. Are there any examples of measuring the difference it makes when those elements are included?有一種說法是:具備真實體驗、以聲音講述、事先準備好信任,會讓人產生被理解的感覺,也更容易形成同意。請問有沒有實例,測量過加入這些要素之後的差異呢?
I cannot claim a definitive comparison study from this session. Lived experience, voice and preparation can make people feel understood, but they do not guarantee agreement. A useful study would compare text, audio and embodied formats, then measure comprehension, willingness to correct the record, who stayed in the process and how views changed later. Trust should be observed, not assumed.
投票信念はいまはイエスの方向かもしれないが、時間が経ったり、別の情報を入れることで安易に変更になることもあるかもしれません。そのようなアンバランスを生まないような工夫はされていますか?The belief behind a vote may point towards yes at the moment, but it might also change all too easily with the passage of time or the introduction of other information. Has anything been devised so as not to produce that kind of imbalance?投票的信念現在或許朝向「Yes」的方向,但隨著時間經過、或接收到其他資訊,也可能輕易改變。請問有沒有什麼設計上的用心,能避免產生這樣的不平衡呢?
Answered已回答回答済みWritten response書面回覆書面回答
Belief drift is not automatically a defect. The main hall discussed today how deliberation can overturn an earlier poll. That change can be a feature when it is transparent: record the starting view, the new information or conversation, the revised view and the person’s reason for changing. Do not freeze the first vote. Make the movement inspectable.
課題感の合意形成と課題解決方法の合意形成など枠を適切に定めてその目的に対して本質的なstatementに制約しないと、全員がやんわり合意できる当たり前な結論しかでない可能性があるのではないか?Unless the frame is properly set, distinguishing consensus on the sense of the problem from consensus on how to solve it, and statements are constrained to those essential to that purpose, isn't there a possibility that the only outcome is an obvious conclusion everyone can mildly agree with?如果沒有適切地界定「針對課題意識的共識」與「針對解決方法的共識」等框架,也沒有針對該目的把陳述(statement)限縮在本質性的內容上,是否可能最後只得出人人都能溫和同意的理所當然結論呢?
Answered已回答回答済みWritten response書面回覆書面回答
Yes. Agreement that something hurts is different from agreement about what to do. Set the scope of the problem first, then ask a separate question about possible remedies. Each phase needs statements that serve its purpose, with the original reasons retained. If both agreements are blurred together, the process can shrink into a bland conclusion that nobody actively opposes.
crowDEA非常に面白かったです。こういうものが当たり前になるといいなと思いますが、逆に不向きなテーマやAIモデル上取り扱いが難しい性質のテーマはあるのでしょうか。crowDEA was very interesting. I hope this sort of thing becomes commonplace; conversely, though, are there themes it is unsuited to, or themes whose nature is difficult to handle on an AI model?crowDEA 非常有趣。我希望這樣的東西能變得理所當然地普及;但反過來說,是否也有不適合的主題,或在 AI 模型上性質難以處理的主題呢?
Answered已回答回答済みWritten response書面回覆書面回答
The product limits and roadmap for crowDEA belong to its makers, so I will not invent them. The principle is clearer: themes built on embodied trust, grief or protected confidences need human-paced rooms. AI mediation is a good first fit where ideas are abundant and stakes are lighter. That is a boundary, not a rejection of the tool.
声が届く、という成功体験をもっと頻繁に小さく得られる社会をどうやって作ればいいのだろう、と考えさせられましたIt set me thinking about how we could build a society where that small experience of success, of one's voice getting through, can be had far more often.這讓我思考:要怎麼打造一個社會,讓「聲音有傳達出去」這種成功經驗,能更頻繁地、以更小的規模被獲得。
Answered已回答回答済みWritten response書面回覆書面回答
That small, frequent experience of being heard is the closed loop this page is trying to demonstrate tonight. Shrink the loop to a classroom or a street, then publish a receipt: whose voice arrived, what changed and who answered. The page you are reading is one worked example. A democracy grows confidence when the return journey is short enough to notice.
示された高校の事例でCrowDEAがしていたことというのは人間のいわゆるファシリテーターの仕事と同様で、それを個人技に頼らずできるということでしょうか。あるいは、それ以上の何かなのでしょうか。In the high-school case shown, is what CrowDEA was doing the same as the work of a so-called human facilitator, meaning it can be done without relying on individual skill? Or is it something beyond that?在所展示的高中案例裡,CrowDEA 所做的事,是否就等同於一般所謂人類引導者(facilitator)的工作,也就是說可以不必依賴個人技藝就能做到?還是說,它是超出那之上的某種東西呢?
Answered已回答回答済みWritten response書面回覆書面回答
CrowDEA can make the repeatable part of facilitation more available: turn-taking, clustering and memory. It does not replace the facilitator’s care, timing or reading of a room. The more human work is noticing hesitation, making space and receiving what has not yet become a sentence. AI is useful when it frees facilitators for that work, rather than pretending to be it.
多元主義的な価値観を持つ人々が独善的に信念を持ちながら、相手を殲滅しようとして分断と抗争が起こる、それをビックテックがSNSで促進しているのが現在だと思います。この大きな波にどう対抗できるでしょうかI think the present situation is that people with pluralistic values hold their beliefs self-righteously and try to annihilate the other side, giving rise to division and strife, and Big Tech is fuelling this on social media. How can we stand up to this great wave?我認為現況是:懷抱多元主義價值觀的人們自以為是地堅持信念,企圖殲滅對方,因而產生分裂與抗爭,而大型科技公司(Big Tech)正在社群媒體上助長這一切。面對這股巨大的浪潮,我們能如何對抗呢?
Answered已回答回答済みWritten response書面回覆書面回答
The pain of division is real, but we should not amplify a frame that puts opponents beyond dialogue. The structural question is how engagement economics profits from the gap between people, not which company becomes our villain. We can answer with prosocial defaults, metrics for bridging statements, interoperability and a genuine right to leave, while keeping offline rooms in the loop.
AIが媒介する際のバイアスに関してご質問したいです。AIのフィルターを通して個人の声がそのまま届くとも限らないケースもあるんじゃないかなと思います。AIの中立性の展望とその限界についてお伺いしたいですI would like to ask about bias when AI mediates. I suspect there are cases where an individual's voice does not necessarily arrive intact once it passes through an AI filter. I would like to hear your outlook on AI neutrality and its limits.想請教關於 AI 居中媒介時的偏誤。我想也會有這樣的情況:經過 AI 的過濾,個人的聲音未必能原封不動地傳達。想請問各位對 AI 中立性的展望及其侷限的看法。
Answered已回答回答済みWritten response書面回覆書面回答
There is no neutral AI. Training data, the designer’s question and the summarising rule all shape what arrives. The remedy is inspectability: open prompts and logs, show originals beside summaries and let people audit the path, as Osugi-san modelled today. The raw voice must also remain available, with a right to bypass the filter whenever mediation changes the meaning.
青山さん、落ち着いて、ちょっとゆっくりはなしてほしい。緊張かなAoyama-san, please relax and speak a little more slowly. Nerves, perhaps?青山先生,放輕鬆,希望您稍微講慢一點。可能是緊張了吧?
Answered已回答回答済みWritten response書面回覆書面回答
Aoyama-san, thank you for supporting a second room after delivering your own tutorial today. There is no need to hurry. The run-of-show design and the organisers should set the pace; it should not be placed on you alone.
それぞれのアプローチから、いまや無視できないレベルになってる社会問題におけるフィルターバブルやそれによる分断をどう解消・抑止していけそうか聞いてみたいですI would like to hear from each of your approaches how the filter bubbles that have now reached a level impossible to ignore as a social problem, and the division they cause, might be resolved or held in check.想聽聽各位分別從自身的取徑出發,認為對於如今已嚴重到無法忽視的社會問題,即過濾氣泡(filter bubble)及其帶來的分裂,可以如何加以化解或遏止。
Answered已回答回答済みWritten response書面回覆書面回答
Broadcasting more of the same will not dissolve a filter bubble. We need to reward statements that people on both sides can endorse, in the Polis lineage, diversify the default feed and keep a route back to offline conversation. Bridging is a design choice, not a mood. Small, visible connections can make division less self-reinforcing.
把同樣的意見播得更多,無法消解過濾氣泡。我們需要獎勵不同立場的人都能認同的句子,延續 Polis 的脈絡,讓預設資訊流更多元,也保留回到線下對話的路。橋接是設計選擇,不是心情好壞。小而可見的連結,能讓分裂不再一再自我強化。
そもそも話す気がない人、話したくない人というのも考えられますが、そのような方々に対して、技術を用いてどのように対応するのでしょうか。仕方ないとしてその人の意見を排除することはすくなくとも政治においてはあってはならないことだと考えています。Presumably there are also people who have no intention of talking, or who do not want to talk. How would technology be used to deal with such people? Shrugging it off as inevitable and excluding their opinions is something that, at least in politics, I believe must never happen.也可以想見有些人根本無意開口、或不想說話。對於這樣的人們,要如何運用技術來因應呢?就算當作沒辦法而把那些人的意見排除在外,至少在政治領域,我認為是絕不該發生的事。
Answered已回答回答済みWritten response書面回覆書面回答
Silence is not consent, and non-participation must not become exclusion. With consent, we can use sortition-based invitations and carefully read ambient or administrative signals, while letting each person choose any proxy who speaks for them. The door should stay open for a late arrival. A listening loop is healthy only when people can enter it, or decline it, on their own terms.
現状AIを通じた自己理解は言語偏重を、ひいては曖昧さを殺さないか? 絵、手触り、握手やハグ、話さないことでの通じ合いを特にAIが身体らしきものを今後得たとしても、他者との触れ合いを阻害することはないだろうか。感情吹き出しチャットにしても、その形に異議申し立てが出てきて、例えば自己の感情に対する自己理解ではないだろうか。As it stands, doesn't self-understanding through AI lean on language and, by extension, kill ambiguity? Pictures, texture, handshakes and hugs, understanding reached through not speaking: even if AI acquires something body-like in future, might it not impede that contact with others? Even with the emotion-bubble chat, objections to that form will arise; for instance, is it not really self-understanding of one's own emotions?現階段透過 AI 的自我理解,是否偏重語言,進而扼殺了曖昧性?圖畫、觸感、握手或擁抱、藉由不說話而達成的心意相通,即使 AI 今後獲得類似身體的東西,會不會反而妨礙了與他者的接觸?就連「情緒泡泡」聊天,對那樣的形式也會出現異議,例如:那難道不是對自身情緒的自我理解嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, language-heavy AI can flatten ambiguity, the pause and the knowledge carried by presence. We do not need to turn drawing, touch, a handshake, a hug or silence into text. An emotional bubble should remain a hypothesis, not a verdict about the self. Good mediation protects wordless channels and knows when measurement must stop.
是的,偏重語言的 AI 可能把曖昧、停頓,以及在場本身傳遞的理解壓平。我們不需要把畫畫、觸碰、握手、擁抱或沉默都轉成文字。情緒泡泡應該只是待驗證的假設,不是對自我的判決。好的媒介設計會守護無言的溝通,也知道何時該停止測量。
異なる意見の合意形成は、他人の自己理解を目撃違いを通じてし合う瞬間に生まれるのに、テクノロジーがその機会を奪うことで各々の孤立を助長しないだろうか。Consensus between differing opinions is born in the moments when people witness one another's self-understanding through their differences, so might technology, by taking those opportunities away, not end up abetting each person's isolation?不同意見之間的共識,本是在人們透過差異、相互見證彼此自我理解的瞬間中誕生的;技術若奪走這樣的機會,難道不會助長每個人的孤立嗎?
Answered已回答回答済みWritten response書面回覆書面回答
That concern matters. A tool that organises voices must not replace the time in which people witness one another’s hesitation, care and changing understanding. Keep the face-to-face encounter, and let technology support the before and after with a visible record. A loop that reduces isolation must return to the space between people, not merely produce a better summary.
プライバシー問題で、他人に話しをするか否かの判断は、本人の意志によるものでなければならないと思います。人ならダメで、AIならOKは違和感があります。人でもAIでも本人が認めた相手とすべきではないのでしょうか?On the privacy issue, I believe the judgement of whether or not to talk to another must rest on the person's own will. 'Not allowed with a human, but fine with an AI' feels off to me. Whether human or AI, should it not be someone the person themselves has accepted?在隱私問題上,我認為要不要向他人訴說,必須取決於當事人自身的意願。「對人不行、對 AI 就可以」讓我覺得不對勁。無論是人還是 AI,難道不都應該以當事人認可的對象為準嗎?
Answered已回答回答済みWritten response書面回覆書面回答
I share that unease. The speaker decides who may receive a voice. Whether the listener is a person or an AI comes second. The purpose, retention and sharing boundary should be clear before anyone speaks, with a genuine choice of recipient and use. Consent is not a one-off permission slip. It must remain reversible when circumstances change.
LLMによる言論仲介や言論抽出を大規模で行う場合、LLM側から人間へのバイアス注入作用が無視できなくなるのではないか?When discourse mediation and discourse extraction by LLMs are done at large scale, won't the injection of bias from the LLM side into humans become impossible to ignore?若由 LLM 大規模進行言論仲介或言論擷取,來自 LLM 一方對人類注入偏誤的作用,是否會變得無法忽視?
Answered已回答回答済みWritten response書面回覆書面回答
At scale, an LLM can quietly steer how people see an issue, and that effect cannot be waved away. The remedy is not to claim neutrality. Keep raw voices beside summaries, make prompts and logs inspectable, compare plural models and preserve a route back for human correction. Mediation earns trust when its influence can be examined rather than hidden.
声なき声を可視化することと、自己を理解することは、裏表なのだろうけれど、そこでテクノロジーがどのように機能するのか、具体的なイメージを持ちきれない。Making voiceless voices visible and understanding oneself are probably two sides of one coin, but I cannot quite form a concrete image of how technology functions there.讓無聲的聲音被看見,與理解自我,兩者大概是一體兩面;但技術在其中究竟如何運作,我還無法形成具體的想像。
Answered已回答回答済みWritten response書面回覆書面回答
Making unheard voices visible can do two things at once: help others notice them and give the speaker a mirror for understanding their own thought. Technology can group related voices, link back to the original and welcome correction by the person who spoke. That return journey matters. Visibility becomes self-understanding through an exchange, not through extraction alone.
民主主義の選挙とかって、「自分の声」がきいている感覚が大事だから、とてもよいとおもうWith democratic elections and the like, the feeling that 'my own voice' is having an effect is what matters, so I think they are a very good thing.民主的選舉這類事情,因為「自己的聲音」有發揮作用的那種感覺很重要,所以我覺得非常好。
Answered已回答回答済みWritten response書面回覆書面回答
Yes, that feeling of one’s own voice being heard is a vital texture of democracy. It should not end on election day. A receipt can show what entered the decision, who answered and what changed next. When everyday settings also offer small choices followed by visible returns, people can feel their voice shaping the shared world between elections.
先ほどから度々「善い」という判断基準がお話の中で出てきますが、みなさんはそれぞれどのようなことを「善い」と考えられていますか?何が善なのか、という価値観が昨今揺らいでいるように思えます。The criterion 'good' has come up repeatedly in the discussion just now. What does each of you consider 'good' to be? The sense of what counts as good seems to have been wavering of late.從剛才起,「善」這個判斷基準在談話中屢屢出現。請問各位分別認為什麼樣的事情是「善」的呢?「何謂善」的價值觀,近來似乎正在動搖。
Answered已回答回答済みWritten response書面回覆書面回答
I cannot appoint one universal Good on everyone’s behalf. People rightly hold several goods, such as safety, freedom and solidarity, and they may pull in different directions. Our task is to build procedures where those goods can coexist, then leave receipts showing who chose what and what changed. When values are unsettled, plural answers need stronger public footing.
珍しい意見をアルゴリズム的にピックアップしても、その意見を発する人の身体性を共有していないと、その意見が理解されず会話の俎上に上がらない問題が限界としてありそうです。(私は意見が理解されないことが多いという悩みがあります)Even if unusual opinions are picked up algorithmically, unless the embodied experience of the person voicing them is shared, those opinions are not understood and never reach the table of conversation; that seems likely to be a limit. (I myself worry that my opinions are often not understood.)即使用演算法把罕見的意見挑出來,若沒有共享發言者的身體性,那個意見仍不會被理解、無法端上對話的檯面,這似乎會是一種侷限。(我自己就有意見常常不被理解的煩惱)
Answered已回答回答済みWritten response書面回覆書面回答
The experience of often having one’s view misunderstood deserves real care. An algorithm can surface a rare contribution, but it cannot supply the speaker’s embodied experience. Give the person time to add context and correct the record, without rushing their life into a category. Technology can widen the entrance, while human listeners make room for the person behind the signal.
馬場先生のイルミディアは、政治や民主主義のような難しいテーマよりも、身近な日常で使えそうなツールとして魅力的でした。日常の場面でデジタル民主主義に近いことを取り入れるために工夫していることがあれば知りたいです。Baba-sensei's Ilmidia appealed to me as a tool usable in familiar, everyday settings rather than for difficult themes such as politics or democracy. I would like to know whether there is anything you deliberately do to bring something close to digital democracy into everyday situations.馬場先生的 Ilmidia,比起政治或民主這類困難的主題,作為能在貼近生活的日常中使用的工具更有魅力。想知道各位為了在日常場景中帶入接近數位民主的做法,有沒有什麼特別的用心。
Answered已回答回答済みWritten response書面回覆書面回答
Ilmidia and crowDEA are attractive because they bring democratic practice down from a grand institution to an everyday choice. A family can use them for plans, a classroom for roles or shared rules. Shape the question together, make options visible and return a small receipt after deciding. Repeated micro-democracy is how participation becomes an ordinary civic muscle.
政治をいかに発酵できるかですよね…It comes down to how we can ferment politics, doesn't it…重點就在於,政治要如何才能發酵吧……
Answered已回答回答済みWritten response書面回覆書面回答
Gō-san, that is a delightful metaphor. Fermentation needs time, a living culture and the right vessel. Politics needs the same: room for different voices, patient attention and a container that can hold change without sealing it away. Small conversations accumulate, and their receipts let us taste how an agreement has matured rather than pretending it appeared fully formed.
合意形成という遅いプロセスと合意形成せずに実行するという早いプロセスがあった場合に往々にして、合意形成せずに実行してしまう人の方が利益を得てしまうという課題もあるように感じた。Given a slow process of consensus-building and a fast process of executing without it, I felt there is also the problem that, as often as not, the ones who go ahead without building consensus are the ones who end up profiting.我感覺還有一個課題:當同時存在「凝聚共識」這種緩慢的過程,與「不經共識就執行」這種快速的過程時,往往是不經共識就執行的人反而得利。
Answered已回答回答済みWritten response書面回覆書面回答
The asymmetry is real: acting without agreement can bring an immediate advantage while those seeking consent carry the delay. A fair design records who accelerated, who benefited and who bore the cost, even when the process moved ahead. Public receipts make the shortcut accountable. Skipping consent should not be free simply because it is fast.
"合意"と個人と集団の"利益"の関係性をどのように整理すると良いのかが非常に難しいと感じた。CrowDEAでは選択の意図と選択結果の利益をどう分けているか。I felt it is very hard to know how best to sort out the relation between 'consensus' and the 'interests' of the individual and the group. In CrowDEA, how are the intention of a choice and the benefit of its outcome separated?我深感困難的是:「共識」與個人及群體的「利益」之間的關係,究竟該如何梳理才好。想請問在 CrowDEA 中,選擇的意圖與選擇結果的利益是如何區分的。
The exact CrowDEA implementation belongs to its makers, so I will not invent a product answer. The principle is to record intention and outcome separately: what people hoped for, who benefited and who carried the cost. That distinction stops consensus becoming a synonym for benefit. A decision trail can then show where the choice needs to be revisited.
人間は性善説だけでなく性悪説で捉える必要もある存在ですが、性悪なふるまいをする人間が場に参加する場合、デジタル技術はそういった人物を排除することになるのでしょうか?Human beings need to be viewed not only through the theory of innate goodness but also that of innate badness. When someone who behaves badly joins a forum, will digital technology end up excluding such a person?人是不僅需要以性善說、也需要以性惡說來看待的存在。當做出惡劣行為的人參與場域時,數位技術是否就會走向排除這樣的人呢?
Answered已回答回答済みWritten response書面回覆書面回答
We should not make entry depend on declaring someone good or bad. Harmful repetition, intimidation or coordinated abuse can meet limits on the behaviour: rate limits, reporting and moderation. Those safeguards should constrain harm, not persons, with a route to appeal and return. The aim is a room that stays open while making it harder for harmful conduct to dominate.
住民か諦めている数々のイシューを可視化し、課題として共有することで、非常に可能性があると感じました。フィードバックのやり方によれは、自治体の窓口を楽にするソリューションにも感じられました。各自治体でも横のやり取りが無く、窓口が疲弊しているのではと想定します。いかがでしょうか。I felt real potential in making visible the many issues residents have given up on and sharing them as challenges. Depending on how the feedback is handled, it also felt like a solution that could make life easier for municipal service counters. I imagine there is no horizontal exchange between municipalities either, and the counters are worn down. What is your view?讓居民已經放棄的眾多議題被看見、並作為課題共享,我覺得非常有潛力。依回饋方式的不同,我也覺得它可以成為讓地方政府窗口更輕鬆的解決方案。我推測各地方政府之間也缺乏橫向交流,窗口早已疲憊不堪。不知道您怎麼看。
Answered已回答回答済みWritten response書面回覆書面回答
Hirota-san’s issue inventory can become shared infrastructure, not another burden placed on municipal counters. If one answer and its change receipt can travel horizontally, staff need not repeat the same explanation in every locality, and residents’ fatigue-driven silence becomes visible. That is the same closed-loop principle as bmi: return an answer once, then let others reuse the receipt.
先ほど挙げていただいた課題の解決のためには、サービスでの解決すべきものと、制度そのものを変える必要があると思いますが、現状は両方行っているのでしょうか??To solve the issues you have just raised, I think some things should be solved through the service while the system itself also needs changing. At present, are you doing both??為了解決您剛才列舉的課題,我認為有應該透過服務解決的部分,也有必要改變制度本身,目前是兩者都在進行嗎??
Answered已回答回答済みWritten response書面回覆書面回答
Both are needed, and sequence matters. Services meet an immediate need and create a receipt showing who answered, when and what changed. When the same gap keeps returning, those receipts make the case for institutional change. Institutional reform without a usable service does not reach people, while a service without reform keeps treating the same symptom. Each gives the other evidence.
ある程度時間のかかる事業をVCから資金調達をして進めるというのは難易度が高そうに感じました。この辺りはどの様な工夫をされていますか?Funding a venture that takes a fair amount of time by raising money from VCs sounds like a high bar to me. What kinds of measures have you devised in this area?向 VC 募資來推進相當花時間的事業,感覺難度很高。這方面你們下了什麼樣的工夫?
The specific financing choices belong to Hirota-san, so I will not guess at them. The wider principle is that civic infrastructure needs patient capital, revenue from government contracts, an open-core model and honest time horizons. A project that grows trust locally cannot be judged only by quarterly expectations. Its financing should give the work enough time to show durable public value.
他のサービスと比較した場合のissuesの特徴として「課題解決」にフォーカスしている点があるように感じました(ほかのサービスは、民主的なプロセスの担保にフォーカスしている印象)。もし、この点について廣田さんのこだわりや、デジタル民主主義の文脈における意義のようなものがあれば、お伺いしてみたいです。Compared with other services, I felt a distinguishing feature of issues is its focus on 'problem-solving' (my impression is that other services focus on guaranteeing democratic process). If Hirota-san has a particular commitment on this point, or sees in it a significance within the context of digital democracy, I would like to ask about it.與其他服務相比,我感覺 issues 的特色在於聚焦「課題解決」(印象中其他服務聚焦於保障民主程序)。如果廣田先生對這一點有什麼堅持,或認為它在數位民主的脈絡中有什麼樣的意義,想向您請教。
The product choices behind issues belong to Hirota-san, but the wider principle joins both aims. We need agreement about what hurts, a visible process for setting priorities and a receipt showing what changed after the proposed solution. Problem-solving gives democracy traction; democratic procedure keeps the solution answerable to the people who described the problem.
市民の声を集めるだけでなく、企業や団体の声も起点にしているところが特徴的だと感じました。今日のイベントの主題はデジタル民主主義ですが、廣田さんからご覧になられて、市民のみならず企業や団体の声も集めることは、デジタル民主主義においてどのような意義や価値があるとお感じでしょうか?I found it distinctive that you start not only from citizens' voices but also from the voices of companies and organisations. Today's event takes digital democracy as its theme: in Hirota-san's view, what significance or value do you feel there is for digital democracy in gathering the voices of companies and organisations as well as citizens?我覺得很有特色的一點是,不只蒐集市民的聲音,也以企業和團體的聲音作為起點。今天活動的主題是數位民主,在廣田先生看來,除了市民之外也蒐集企業、團體的聲音,對數位民主而言具有什麼樣的意義或價值呢?
Answered已回答回答済みWritten response書面回覆書面回答
Listening to companies and organisations can widen the map of people affected, and can be part of broad listening. The safeguard is upstream agreement: say whose issue is being addressed and why before the microphone, and prevent well-resourced voices from automatically moving first. Set those voices beside residents’ accounts, then publish the use and outcome. That is what makes the listening public.
国会議員マップは収支は今も赤字でしょうか?その場合、このサービスを今後継続していくためにどのように運営資金を得ていくか、お考えはありますか?Is the Kokkai Giin Map still running at a deficit? If so, do you have thoughts on how to raise the operating funds to keep the service going?國會議員地圖目前的收支仍是赤字嗎?若是如此,為了讓這項服務持續下去,您對今後如何取得營運資金有什麼想法嗎?
The current balance and funding plan belong to Nakajima-san to answer. In principle, civic infrastructure needs boring, durable money: memberships, grants and public contracts, rather than a single heroic sponsor. An open-source mirror is continuity insurance too, so the map can outlive any one host, budget or moment of attention.
中立を保とうとされているのと、間口を広げようと思ってられるのがとても伝わりました 公開されてからバズるまでどのような経緯がありましたか?It came across very clearly that you are trying to stay neutral and to keep the entrance wide. What was the path from release to the map going viral?非常能感受到您想保持中立、也想拓寬入口的用心。從公開到爆紅,中間經歷了什麼樣的過程呢?
The story from publication to becoming widely shared is Nakajima-san’s to tell. For me, neutrality is not a hidden opinion or a claim of colourlessness. It is an inspectable method: sources, definitions and update history are visible enough for people with different starting points to check. The wider the entrance, the more that matters.
国会議員マップは高機能かつデザインも綺麗で、これを作り切るには相当な根気(または投資)が必要だったと拝察します。中島さんが国会議員マップを作り切ることを可能にした原動力は何なのでしょうか?ちょっとした疑問や思いつきでできることではないと感じます。The Kokkai Giin Map is highly capable and beautifully designed; I imagine carrying it through to completion took considerable perseverance (or investment). What was the driving force that enabled Nakajima-san to see the map through to the end? It does not feel like something a passing question or whim could achieve.國會議員地圖功能強大、設計又漂亮,想必要把它完整做出來,需要相當的毅力(或投資)。讓中島先生得以把國會議員地圖做到完成的原動力是什麼呢?我覺得這不是一點小疑問或一時興起就能辦到的事。
The reason Nakajima-san finished it is his story to tell. Still, what sustains a builder is often not completion alone, but the moment someone uses the work and says, “Now I can see.” When the tools barrier falls, public goods are no longer reserved for people who already know how to code. That is a civic-tech spirit worth celebrating.
法を犯した過去とこういったサービスを作ったこととは関係ありますか?その辺りも意識して作ったのでしょうか。Is there a connection between having broken the law in the past and having built a service like this? Was that something you were also conscious of as you built it?曾經觸法的過去,與打造出這樣的服務之間有關聯嗎?您在製作時是否也意識到了這一點呢?
I have not verified the facts this question presupposes, and I will not address them here. What Nakajima-san chooses to say about himself belongs to his dignity and agency. I can answer only about the work. The Parliament Members Map is verifiable by anyone through its sources, criteria and update history. What we should question is not a biography, but the receipt.
What Nakajima-san wishes to build next is his answer to give. My small wish is that, whatever the subject, it ships with a receipt: a person’s response and a visible sign of what changed. Even a modest receipt can guide the next question, turning a finished tool into a living civic conversation.
貧乏でパソコンも持ってない、パソコン買ったところで自分には何ができるんだ・使いこなせいと思っていたけど、中島さんのお話を聞いて一番大事なのは喧嘩しながらもしぶとく対話することなんだと思いました。ありがとうございます。それができるのがデジタルでありAIなんだと。I am poor and do not even own a computer, and I used to think that even if I bought one, what could someone like me do with it; I would never master it. But hearing Nakajima-san’s talk, I came to feel that what matters most is to keep up a dogged dialogue even while quarrelling. Thank you. And that it is digital and AI that make this possible.我很窮,連電腦都沒有,本來覺得就算買了電腦,自己又能做什麼呢、恐怕也無法上手;但聽了中島先生的分享,我體會到最重要的,是即使會吵架,也要不屈不撓地持續對話。謝謝您。而讓這件事成為可能的,正是數位與 AI。
Answered已回答回答済みWritten response書面回覆書面回答
I am glad this reached us. Before having the right computer, or confidence that we can master it, there is the stubborn civic muscle of continuing the conversation, even through disagreement. Nakajima-san’s story makes that muscle an entrance to digital tools and AI, rather than a qualification we must already possess. Thank you for putting it so clearly.
今天能在這裡收到這句話,我很高興。在擁有合適的電腦、或確信自己能熟練使用之前,我們先有一種堅持:即使爭吵,也繼續對話。中島さん的分享,讓這股力量成為走向數位與 AI 的入口,而不是必須先具備的資格。謝謝你說得這麼清楚。
こんなものが欲しいな、というものに既に追いついているとして、さらに探っていくような機会はまだ残っているのでしょうか。Supposing the map has already caught up with the “I wish something like this existed” wishes, is there still room left to keep exploring further?假設它已經趕上了「真希望有這種東西」的期待,那麼是否還留有可以繼續探索下去的機會呢?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, there is still room to explore. A wish becomes more precise after people use the thing: new frictions appear, unexpected uses emerge and the next question comes back as a receipt. For the parliament map, its users can help reveal that next layer. Wants are not a finish line; they are invitations to keep listening.
ラーニングシティ、とてもいいと思います。が、行政職員のAI利用はガチガチに守られた環境ですごく保守的な利用にとどまっているように思います。どうすればラーニング志向の利用が進むでしょうかThe Learning City: I think it is very good. But AI use by administrative staff seems to stay very conservative, inside rigidly guarded environments. How can learning-oriented use be advanced?學習城市(Learning City),我覺得非常好。不過,行政職員的 AI 使用,似乎停留在被層層防護的環境中、非常保守的用法。要怎麼做,學習導向的使用才會有進展呢?
Answered已回答回答済みWritten response書面回覆書面回答
Guarded environments become conservative when officials fear that a failed experiment will be punished. Give them small sandboxes and leave a receipt: what was entered, who checked it and what changed. That creates a learning cycle where failure is not punished. Osugi-san’s audit-log posture today points to the safer choice: make experiments inspectable rather than locking them away.
東京都のように体力がある自治体だからこそできる話で、明らかに財政力が背景にあるわけです。良い取り組みだとは思いますが、地方出身者からすると複雑な気持ちです。水道や私立高校を無償化できるような東京都ならできることでも、Learningしても財源がなくて何もできないという自治体も多いと思います。This is a story possible precisely because the municipality has Tokyo's stamina; fiscal strength is clearly what lies behind it. I do think it is a good initiative, but as someone from the regions I have mixed feelings. Even where a Tokyo that can make water and private high schools free can act, I think there are many municipalities that, however much they Learn, have no funds and can do nothing.這正是像東京都這樣有體力的自治體才辦得到的事,背後顯然是財政實力。我認為是很好的嘗試,但對出身地方的人而言,心情很複雜。即使是能讓自來水與私立高中免費的東京都辦得到的事,我想也有許多自治體就算做了 Learning,也因為沒有財源而什麼都做不了。
Answered已回答回答済みWritten response書面回覆書面回答
That complicated feeling is entirely justified: Tokyo’s fiscal capacity cannot be treated as a universal starting point. If learning is to be fair, learning itself must be cheap. Open source, shared procurement and the packaged deliberation described today can help poorer municipalities inherit tools, not licences. A receipt need not consume a budget. It can begin as one public webpage.
学習には痛みが必要と思うのですが、誰がどのように痛むのが正しい学習サイクルてしょうか?I think learning requires pain, but who should hurt, and in what way, for the learning cycle to be right?我認為學習需要伴隨痛楚,但由誰、以什麼方式承受這份痛,才是正確的學習循環呢?
Answered已回答回答済みWritten response書面回覆書面回答
This question entered my closing speech minutes after it was asked. It became the night’s fastest closed loop. The pain of learning should be training soreness, not injury. Learning’s burden should be taken up alongside people already carrying the pain, never dumped on whoever is farthest from the microphone. That practice builds civic muscle: the strength to learn together without hiding who pays the cost.
私もパブリックセクターで働いているので、たいへん参考になります。私は、基本的に単年度予算の行政において「施策のKPI設定がそもそも間違っている」と思う場面が多々あるのですが、宮坂さんはKPI設定についてどのように思われますでしょうか?I too work in the public sector, so this is very instructive. Under government's essentially single-year budgets, there are many moments when I think the KPI-setting for a measure is wrong to begin with. Miyasaka-san, how do you think about KPI setting?我自己也在公部門工作,因此非常受用。在基本上採單年度預算的行政體系裡,我常常遇到覺得「政策的 KPI 設定打從根本就錯了」的場面。想請問宮坂先生對 KPI 設定怎麼看呢?
Answered已回答回答済みWritten response書面回覆書面回答
Miyasaka-san’s own view on KPI design belongs to him, but my principle is clear. Single-year budgets often reward visible activity, while the learning that matters arrives between years. Count closed loops, not only meetings or users: what was heard, what changed and what was repaired after a failed trial. A receipt makes failure survivable because the next budget can fund documented improvement.
行政の場合、為政者の判断が正解だったりするので、判断基準は為政者とする、これが正解かどうかは選挙での判断。ということになるのですかね?In government, the judgement of those who govern can turn out to be the right answer, so the criterion of judgement rests with them, and whether that was right is judged at elections. Is that what it comes down to, I wonder?就行政而言,有時執政者的判斷就是正解,所以判斷基準以執政者為準;至於這是否為正解,則交由選舉來判斷。事情就會變成這樣嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Elections are one slow learning system: people review a government’s choices together and render a judgement. Daily receipts do not replace that electoral judgement. They complement it by showing, in real time, what was heard, who decided and what changed. Small closed loops give people better material for the larger review that arrives at election time.
東京都が学んだ結果をAI を通して地方自治体に配布し、小さい調整サイクルだけで実行できるようになると良い効果的なラーニングをできるのではないかと感じています(FTのイメージ)。地方自治体を横断するラーニングをどうすればできるか、できる領域とできない領域は何かなど目星はあるでしょうかI feel that if the results of Tokyo's learning were distributed to local municipalities through AI, and they could execute with only small cycles of adjustment, effective learning might be possible (something like FT is the image). Do you have any sense of how learning across municipalities could be done, and of the areas where it can and cannot work?我覺得,若能把東京都學習的成果透過 AI 發布給地方自治體,讓各地只需小幅的調整循環就能執行,或許就能達成有效的學習(類似 FT 的想像)。關於跨自治體的學習要怎麼做才可行、哪些領域做得到、哪些做不到,各位心中是否已有眉目呢?
Answered已回答回答済みWritten response書面回覆書面回答
Cross-municipal learning works when patterns and code can travel, but trust is grown locally. Share open models and practical playbooks, then let each local facilitator adapt them to the place and leave a receipt of what changed. The aim is not to distribute one central answer. It is to let municipalities compare closed loops and improve together.
AIが東京のことを深く正しく学習するために、ぜひ、質の高いオープンデータを、よりたくさん公開してくださいSo that AI can learn about Tokyo deeply and correctly, please do release more high-quality open data.為了讓 AI 能深入且正確地學習關於東京的事,請務必公開更多、更高品質的開放資料。
Answered已回答回答済みWritten response書面回覆書面回答
Yes, please, open data. It is how a city lets another city learn from its experience rather than start from zero. Publish it in reusable form, with its purpose, update date and limits. Then the data becomes more than a display. It becomes a receipt that another place can test, adapt and return.
テクノロジーでラーニングしてそれで行政が良くなるというのは水道無償化さえできる財政力が背景にある東京だからできる話で、東京都のように体力がない自治体ではラーニングしても財源がなくて何もできないということはあると思います。民主主義という側面を考えると本当にその前提となる平等性が地域間で確保できているのか本当に疑問です。Learning through technology and thereby improving government is a story possible for Tokyo, backed by fiscal strength that can even make water free; in municipalities without Tokyo's stamina, I think it can happen that they learn yet, having no funds, can do nothing. Viewed from the side of democracy, I genuinely doubt whether the equality it presupposes is being secured between regions.用科技來學習、進而讓行政變好,這是有著連自來水都能免費化的財政實力作為後盾的東京才辦得到的事;像不具東京都那種體力的自治體,我想就會發生即使學習了、也因沒有財源而什麼都做不了的情況。從民主這個面向來想,作為其前提的平等性是否真的在地域之間獲得確保,我實在感到懷疑。
Answered已回答回答済みWritten response書面回覆書面回答
Regional inequality must not be misread as a difference in motivation. That is why the learning loop itself has to be cheap. Shared tools and public playbooks let each place run a small, inspectable experiment. The receipt need not be an expensive system. A public webpage showing what was tried and what changed can already make learning travel.
感想です。ラーニングシティなのですが、正直、都民税は議員の方が学ぶためにお金は払ってませんし、生きていません。過程には必須だと思います。それより、テクノロジーによって質の高い実行力を目指す、という方がはるかに納得感が高いです。This is a comment. About the Learning City: honestly, Tokyo residents' tax is not being paid so that assembly members can learn, nor is it being put to good use. For the process, I think it is essential. But aiming for high-quality execution through technology is far more persuasive to me.這是感想。關於學習城市,老實說,都民稅並不是為了讓議員們學習而繳的,錢也沒有被活用。我認為它對過程而言是必要的。但比起這個,「以科技追求高品質的執行力」的說法,遠遠更能讓人信服。
Answered已回答回答済みWritten response書面回覆書面回答
Thank you for putting that so plainly. Technology and learning matter in the process, but for people paying Tokyo resident tax, execution is the receipt that makes the work credible. If learning never lands in a decision or a working institution, it becomes tuition without a diploma. We should record what was tried, who decided and what changed, then make implementation visible.
東京アプリはマイナンバーカードが必須でトラストがないです。聞くだけだと一方通行で都政への信頼は育ちませんよね。The Tokyo app makes the My Number card compulsory, and there is no trust. Just listening is one-way traffic; trust in the metropolitan government will not grow that way, will it?東京 App 必須有 My Number 卡才能使用,毫無信任可言。只是聽取的話就是單向通行,對都政的信賴是培養不起來的吧。
Answered已回答回答済みWritten response書面回覆書面回答
If the My Number card is a mandatory gateway, saying 'we are listening' cannot by itself create trust. The app's detailed policy belongs to its operators, but the principle is clear: explain the purpose, data retained, alternatives and what changed because people spoke. People without that card must have an equally real route into the loop. Listening becomes democratic only when it returns a visible receipt.
如果以 My Number 卡作為必經入口,只說「我們正在聆聽」本身不會產生信任。應用程式的具體政策要由營運方回答,但原則很清楚:說明用途、保存哪些資料、有哪些替代方式,以及哪些改變來自民眾的聲音。沒有這張卡的人,也必須有同樣真實的參與路徑。聆聽要回到一張可見的收據,才會成為民主的互動。
日本の民主主義の議論は国政中心になりがちな傾向があったと思いますが、テクノロジーを通じて「地方自治は民主主義の学校」的なものが生じると考えてよいでしょうか。Debate about democracy in Japan has tended, I think, to centre on national politics. May we take it that through technology something like 'local self-government as the school of democracy' will emerge?我想日本關於民主的討論,過去往往有以國政為中心的傾向;那麼,可以認為透過科技,將會出現「地方自治是民主的學校」這樣的現象嗎?
Answered已回答回答済みWritten response書面回覆書面回答
Yes, local self-government can become a school of democracy, when technology supports a two-way practice rather than delivering answers from above. People need to frame questions, hear disagreement, inspect the record and receive a visible account of what followed. Each small loop teaches democratic agency. Technology is not the teacher; it is the infrastructure that helps the lesson travel and return.
行政が「奪う」ことができるのはそうですが、近年ではプラットフォームのモデレーションや優先表示のアルゴリズムのような権力もあるわけで、そういった者にまで立憲的な価値観を適用・応用していけないかというのがデジタル立憲主義の議論という理解。It is true that government can 'take away'; but in recent years there are also powers such as platform moderation and ranking algorithms, and my understanding is that digital constitutionalism is the debate about whether constitutional values can be applied and extended even to such actors.行政能夠「剝奪」確實如此,但近年來也存在平台的內容審核或優先顯示演算法這類權力;我的理解是,數位立憲主義的討論,正是在問能否把立憲的價值觀也適用、延伸到這類主體之上。
Answered已回答回答済みWritten response書面回覆書面回答
Yes, that is a useful reading of digital constitutionalism. When moderation or ranking shapes whose speech is seen and heard, it exercises a form of power even when the actor is a private platform. Safeguards should include stated purposes and standards, reasons for decisions, routes to challenge them and meaningful review. Constitutional values belong in digital design wherever concentrated influence can shape people’s lives.
財政力があるからこそ、是非、国と張り合うのではなく、自治体はこう言うことを求めているんだと、実際に都民の窓口となる区市町村のためのデジタル政策をお願いします。Precisely because you have the fiscal strength: please, rather than squaring off against the national government, give us digital policy for the wards and municipalities that actually serve as the point of contact for Tokyo residents, showing that this is what local governments are asking for.正因為有財政實力,才更希望不是與國家較勁,而是表明「自治體所要求的是這樣的東西」,請為實際作為都民窗口的區市町村,推動數位政策。
Answered已回答回答済みWritten response書面回覆書面回答
That is the right test for Tokyo's fiscal capacity: not whether it can compete with the national government, but whether ward and city offices can serve people more clearly. Specific policies belong to their owners to answer. In principle, Tokyo's resources should support shared foundations, reusable components and small experiments shaped with local staff. The loop closes when municipalities can say what they needed and what changed.
マイナンバー配布と同じように、アプリ入れるために、お金をばら撒くのは違うと思いますScattering money about to get the app installed, just as with the My Number rollout, is wrong, I think.就像發放 My Number 時那樣,為了讓人安裝 App 而撒錢,我認為是不對的。
Answered已回答回答済みWritten response書面回覆書面回答
That discomfort matters. Paying for installs may increase a number, but it cannot purchase understanding of the purpose, confidence about data use or a reason to keep participating. The policy details belong to its owners. The principle is that participation should remain a choice, people without the card need another route and public money should be judged by what changed, not by downloads alone.
東京アプリの基盤の上でXXX区など特定の自治体を対象にその役所が調査するとか周知するとかそういう機能があると良いですね。別途でやるのは二重投資で利便性も落ちるのでIt would be good to have functions on top of the Tokyo app's foundation whereby, for a particular municipality such as XXX Ward, its office could run surveys and put out notices. Doing that separately would be double investment and less convenient.若能在東京 App 的基礎之上,針對 XXX 區等特定自治體,提供讓該公所進行調查、宣導公告之類的功能,那就好了。因為另外單獨做既是雙重投資,便利性也會下降。
Answered已回答回答済みWritten response書面回覆書面回答
The Tokyo app’s roadmap belongs to its responsible team, so I will not promise a feature. The principle is sound: a shared foundation could let each ward or municipality run its own surveys and notices without rebuilding the same doorway. That only helps if the purpose, data handling and non-app route are clear, and local offices retain responsibility for the decisions that follow.
サービス授受の対象性についですが、税金を納めることと、サービスを考えることがコンフリしているのが現状かと思いますが、そちらについて感想を伺えますか?About the symmetry of giving and receiving services: my sense of the present situation is that paying taxes and thinking about services are in conflict. Could I hear your reflections on that?關於服務授受的對稱性,我想現況是「繳納稅金」與「思考服務」彼此衝突。能否請教您對此的感想?
Answered已回答回答済みWritten response書面回覆書面回答
I would be careful about treating tax and public service as a one-to-one exchange. That leaves out people who have less time, money or voice, as well as those who do not yet use a service. Tax sustains shared ground; public service returns as changed conditions of life. The democratic test is a visible loop: what was heard, what changed and what decision follows.
市民参加型により市民が何かしらの活動を行ったり事業を行うことは町内会・自治会が負っていた役割のように思います。そうした昔からの活動が見直されてもよいかもしれませんね。Citizens doing some kind of activity or undertaking through citizen participation seems to me the role that neighbourhood and residents' associations used to carry. Perhaps such long-standing activities deserve to be looked at afresh.透過市民參與,讓市民進行某些活動或事業,我覺得這本是町內會、自治會所承擔的角色。這類行之有年的活動,或許也值得重新受到重視呢。
Answered已回答回答済みWritten response書面回覆書面回答
Neighbourhood and residents’ associations have long carried practical knowledge about noticing local needs and acting together. That knowledge deserves renewal, not simple repetition. Participation should welcome people with different amounts of time and different ways of joining, while respecting the choice not to join. Technology can make handover easier without replacing relationships, creating a civic loop more people can enter.
わかりやすい見た目でなく、本当に、都民や職員に役立つデジタル政策をぜひお願いします。ちょっと政治的なアピールみたいなことだらけのでない、何かできませんか?Please give us digital policy that truly serves Tokyo residents and staff, not just an accessible appearance. Could something be done that is not full of what feels like political self-promotion?請務必推出不是只有易懂外觀、而是真正對都民與職員有幫助的數位政策。能不能做點什麼,不要盡是帶有政治宣傳意味的東西呢?
Answered已回答回答済みWritten response書面回覆書面回答
I cannot promise a particular Tokyo policy, but I share the test behind this concern: usefulness matters more than a polished appearance. Ask residents and staff where work is difficult, measure what becomes easier and publish what changed, including who was not served by the digital route. Trust grows from improvements that remain useful after the announcement.
東京都であれば人材が豊富なので、交付税制度を見直し、本来の都の税での地方自治での運営は可能と思います。しかしながら、地方の町村は人材不足でもし金があったとしても運営が困難になってくると思います。ある程度、国や都道府県での行政事務の集約は必要ではないでしょうか?どちらかいうと東京都が特別だと思います。Tokyo has abundant personnel, so I think that if the local allocation tax system were reviewed, running local government on the metropolis's own taxes, as originally meant, would be feasible. Regional towns and villages, however, are short of people, and even with money I think their administration will grow difficult. Isn't some degree of consolidation of administrative work at the national or prefectural level necessary? If anything, I think it is Tokyo that is the special case.東京都的話人才充沛,所以我認為只要重新檢視地方交付稅制度,以都原本的稅收來營運地方自治是可能的。然而地方的町村人才不足,就算有錢,營運也會愈來愈困難。某種程度上把行政事務集約到國家或都道府縣,難道不是必要的嗎?要說的話,我認為東京都才是特別的。
Answered已回答回答済みWritten response書面回覆書面回答
That distinction matters. Devolution cannot mean moving funds while ignoring the uneven distribution of people and skills. Common administrative infrastructure and routine work can be shared across national or prefectural levels, while local priorities and final judgement remain close to the people affected. Tokyo’s method should not be copied wholesale. What can travel are open tools, training and operating patterns, adapted by each municipality to its own conditions.
失敗から学ぶは確かに。失敗したものを失敗しましたということから、新たなものが生まれるかとLearning from failure, certainly. From saying of what failed, 'it failed', something new may be born, I'd say.從失敗中學習,的確如此。願意對失敗的事說出「失敗了」,或許新的東西就會從中誕生吧。
Answered已回答回答済みWritten response書面回覆書面回答
Yes. Naming a failure honestly can be the beginning of a new attempt, but admission alone is not learning. We need to record what was tried, which assumption failed, who carried the cost and what small change will be tested next. When people can inspect the process without being blamed, failure becomes shared public knowledge rather than a hidden reason to avoid experimenting.
Pollの質問に真正面から答えてない。分権化すれば東京に財源が集中するのは確実。地方は目の前にやることがあるのに財源や人手の問題でできないのが課題なのだから、答えになってない。This does not answer the Poll question head-on. Decentralise, and it is certain that funds will concentrate in Tokyo. The regions' problem is that the work is right in front of them and they cannot do it for lack of funds and hands, so this is no answer.這並沒有正面回答 Poll 的問題。一旦分權化,財源集中到東京是必然的。地方明明眼前就有該做的事,卻因財源與人手的問題做不到,這才是課題,所以這樣不算回答。
Answered已回答回答済みWritten response書面回覆書面回答
The call for a direct answer is fair. If decentralisation moves responsibility without reliable funding and people, Tokyo can gain resources while local governments are left with urgent work they cannot deliver. That risk is real. The answer is not simply decentralise. It is to pair local authority with fiscal equalisation, shared specialist capacity and a clear local say over priorities. Those conditions should be explicit before celebrating the transfer of power.
アプリを定期的に習慣的に開き続けるインセンティブが弱い気もしますI also feel the incentive to keep opening the app regularly, as a habit, is weak.我也覺得,讓人定期、習慣性地持續打開 App 的誘因偏弱。
Answered已回答回答済みWritten response書面回覆書面回答
That is an important warning. Increasing the number of app opens should not become the goal, or participation turns into another chore. A stronger incentive is a visible return: what a person’s contribution was used for and what changed afterwards. People should be able to join when it matters, with little friction and a clear way to leave. Habit grows from knowing that a voice reached a decision, not from points alone.
マイナンバーカードで個人認証をしつつも、匿名で個人名は明かされない仕組みがあればいいかと。しっかり実名でいう意見と、匿名でないといえない意見の両者が大切かとIt would be good, I think, to have a mechanism that authenticates identity with the My Number card while remaining anonymous, with no personal name revealed. Both matter, I think: opinions stated firmly under a real name, and opinions that can only be voiced anonymously.我想,若能有一種機制,既以 My Number 卡進行個人認證,又能匿名、不揭露個人姓名,那就好了。堂堂以實名表達的意見,與不匿名就說不出口的意見,兩者應該都很重要。
Answered已回答回答済みWritten response書面回覆書面回答
Identity verification and public anonymity can coexist. A My Number card could be used only to confirm eligibility or prevent duplicate submissions, while the public record shows a separate identifier. We should preserve both accountable, named participation and anonymous participation for voices that need protection. Who can re-identify someone, how long data is kept and when it may be shared must be declared first, with a firm legal boundary against casual disclosure.
身分驗證與公開匿名可以並存。My Number 卡只用於確認參與資格或防止重複投稿,公開紀錄則顯示另一組識別碼。能以實名承擔責任的發言與需要保護的匿名發言都應保留;誰能重新識別、資料保存多久、何時可提供第三方,都要事先說清楚,並以法律程序守住揭露界線。
学び続ける都市全体の力を上げることで、適応や創造性を高めていく大局的な狙いだと思いますが、それを促進する仕掛けの構想はありますか?I take the broad aim to be raising the whole city's capacity to keep learning, and so enhancing adaptation and creativity. Is there a design in mind for mechanisms to encourage that?我想大方向的目標,是藉由提升整座城市持續學習的力量,來提高適應力與創造性。請問是否已有促進這件事的機制構想呢?
Answered已回答回答済みWritten response書面回覆書面回答
The mechanism I would favour is a learning loop, not a single citywide curriculum. Schools, libraries, workplaces and neighbourhood groups can set local questions, run small experiments and publish what they learned and what changed. Others can adapt the evidence, while participants receive a clear reply. Adaptability and creativity grow when the city repeatedly connects learning to decisions, rather than treating learning as a separate service.
AIにすべての行政データを学習させると、AIによるマイナンバーを使ったプロファイリングが容易に実現できてしまいます。このような規制と推進のバランスをどのように考えますか?If AI is made to learn all administrative data, profiling by AI using My Number becomes all too easy to achieve. How do you think about the balance between regulation and promotion here?若讓 AI 學習所有行政資料,AI 利用 My Number 進行側寫(profiling)就會輕易成真。對於這樣的管制與推動之間的平衡,您是怎麼思考的呢?
Answered已回答回答済みWritten response書面回覆書面回答
Regulation and innovation are not opposites. The starting boundary is that government data should never be poured wholesale into one AI system, with a My Number identifier becoming a cross-purpose profile key. Use purpose-specific separation and the minimum necessary data. Anonymisation is not enough: add re-identification testing, staff accountability, audit logs and a way to challenge use. Promote public-interest applications whose boundaries can be inspected, not data concentration.
規範與創新不是二選一。起點是不能把所有行政資料整批匯入同一個 AI 系統,也不能讓 My Number 成為跨目的建立個人檔案的通用鍵。應按目的分隔資料,只取必要部分;匿名化也不夠,還要做重新識別測試、保留職員的最終判斷、留下可稽核紀錄,並提供申訴途徑。推動的是邊界可檢驗的公共利益應用,不是資料集中。
東京アプリでは意見聴取は執行部側がするものなのか、議会がするものなのかの棲み分けや共有はどのようにされていますか?In the Tokyo app, how are the roles divided and shared as to whether gathering opinions is something the executive does or the assembly does?在東京 App 中,聽取意見是由執行部門來做,還是由議會來做,這樣的分工與共享是如何安排的呢?
Answered已回答回答済みWritten response書面回覆書面回答
I cannot confirm the Tokyo app's current division of responsibility without its published operating specification. In principle, the executive side implements services, while the assembly deliberates and scrutinises. Each listening exercise should show who is asking, when and which decision the input will inform. A shared public receipt of responses and follow-up can keep the voice connected even when responsibility is divided.
A closed loop means the answer returns to the people who spoke. You spoke; this page is the return. If an answer here is wrong or too small, reply to it. The loop stays open for corrections, and corrections are also receipts.閉環的意思是:答案回到發言的人手上。你們發了言,這一頁就是回返。如果這裡的答案有錯、或太小,請回信;為了訂正,這個環保持打開,訂正也是收據。閉じたループとは、答えが声を上げた人のもとへ返ることです。あなたが声を上げ、このページがその返答です。もしここにある答えが間違っていたり、小さすぎたりしたら、どうか返信してください。訂正のためにループは開いたままです。訂正もまた、レシートです。
One request, offered as a neighbour: within one year, publish one Japanese closed loop with its receipt, and send the link. It will be retold in Taipei, with your names on it.以鄰人身分的一個請求:一年之內,發表一個日本的閉環,附上收據,把連結寄來。我會在臺北把這個故事重講一遍,署你們的名字。隣人としてのお願いをひとつ。一年以内に、日本の閉じたループを一つ、レシート付きで公開して、リンクを送ってください。台北でその物語を語り直します。名前は、あなたたちのもので。
Licence and provenance: Audrey Tang's closing remarks, answers, and this page structure are CC0. Question texts and Kurimoto's written answers remain their authors' rights and are quoted with source attribution. For a correction, anonymisation, or removal, reply through the channel that shared this page.授權與出處:唐鳳的閉幕致詞、回答與本頁結構採 CC0。問題文字與栗本先生的書面回答仍屬其作者權利,本頁標明出處引用。如需更正、匿名化或移除,請透過分享本頁的頻道回覆。ライセンスと出典:オードリー・タンの閉幕の言葉、回答、ページ構造はCC0です。質問文と栗本さんの書面回答は著作者の権利にとどまり、出典を示して引用しています。訂正・匿名化・削除を希望する場合は、このページを共有したチャンネルを通じて返信してください。
Session quotations are provisional hearing notes until recordings publish; timecode links will then be added; named speakers' corrections take precedence.各場次引文在錄音發布前均屬暫定的聽講筆記;錄音發布後會補上時間碼連結;具名講者的更正優先。セッションの引用は録音公開まで暫定的な聴取メモです。録音が公開されたらタイムコードのリンクを追加します。名前のある話者による訂正を優先します。
Japanese is the source of record for question text and answers; English and Traditional Mandarin are reference translations; translation corrections are receipts.問題文字與回答以日文版為正本;英文與華文是參考譯文;翻譯更正也是收據。質問文と回答は日本語版を正本とし、英語と台湾華語は参考訳です。翻訳の訂正もレシートです。