takashi ikegami

最終講義やります。 池上高志「生命とはなにか？ ～ALIFEの研究から～」 　　  日時：2026年3月13日（金）16：00-17：30 　　  会場：東京大学駒場Iキャンパス 18号館1Fホール u-tokyo.ac.jp/campusmap/cam0…

東京大学共創ラボオープン作戦会議　池上高志　茂木健一郎　第２回　スケール仮説とデータ生成 @alltbl　 ２０２６年５月２６日 #茂木健一郎 #脳の教養チャンネル #東京大学共創ラボ

結局は自民党に3分の2以上の議席を与えたことが原因であって、国家が危機的な状況に置かれているなかこんな荒唐無稽な議論見せられる国民はたまったものではないけど、有権者が学習しない限り同じことやこれ以上酷いことが繰り返されるのは必至であって、まあこのまま落ちていく一方なのでしょう。

Never confuse education with intelligence, you can have a PhD and still be an idiot. - Richard Feynman

いまどき何の意味があるんだろう？　印鑑を押す？なんのために？　１日疲れてこれがトドメかよ。

いまどき請求書に、印鑑を押せとか言われて、呆然としている。印鑑？？？？　印鑑？？？？？？？

@Nonble3 セルフの確立が前提で、内部的に生成されるものなので、本当は外部記述じゃダメですよね。

@alltbl desire は外部観察者による機能的記述で足りるのか、それともエージェント自身にとって顕在している必要があるのか、という区別は重要だと思われますか？

始まりそうだね。

ついに、OpenAIが保有する未公開の汎用モデルが数学の主要な未解決問題を初めて解決 フィールズ賞受賞者ガワーズ氏 『もし人間がこの論文を書いて「Annals of Mathematics」（世界最高峰の数学誌）に投稿し、私が簡単な意見を求められたら、一切の躊躇なく受理（アクセプト）を推奨したでしょう』

これきょうの１８時からだけど、、、荒れそうだなｗ

ずっと迷ってたけど、Gmail読むのは、手伝ってもらうことにした。

The University of Hong Kong, yaşayan dokuyu taklit eden ve canlı hücrelerle etkileşime girebilen hidrojel tabanlı, yumuşak, 3 boyutlu ve biyouyumlu ilk yarı iletkeni geliştirdi.

@Nonble3 身体性の話は大事です。欲求モジュールが大事で、それは身体性にある。

@alltbl 話の全ては理解できてないのですが、なぜ身体性の話が出てこないのでしょうか。 僕はLLMにはエピソード記憶がなくて、エピソード記憶を支えるのは身体性かなと思うのですが。

久しぶりに、新津保さんと、楽しかった。来てくださった皆さん、ありがとうございました。それにしても、酷暑が見える京都でした。

長かったKYOTOGRAPHIE 2026の会期も、いよいよ明日までとなります。最終日となる5/17（日）は、池上高志さんとの、「写真は誰のものか―予測するAIと再配置される主体」と題したパネルディスカッションがあります。kyotographie.jp/events/2026/17…

なにか不思議な時間だった。

稲見さんと、未来の生命について語ることに。久しぶりにお会いできるので楽しみです。

@kenmogi Yet it also creates a somewhat ominous impression, as though an unseen enemy were lurking behind the text. Before turning AI into a political or ideological issue, we should first understand its technical foundations and carefully consider the value of the technology itself.

@kenmogi このレビューは格調が高いし、問題意識も鋭い。

The restless Prometheus: The Technological Republic by Karp and Zamiska. Reading The Technological Republic by Alexander C. Karp and Nicholas W. Zamiska left me deeply disturbed and uneasy. While I recognized the phenomena and symptoms described in the book to be certainly legit, I experienced a surge of restlessness of reason about the wider connotations, after reading the book with pleasure. The authors are spot on in identifying the consumerism behind the recent boom of AI. It may even be acknowledged that there is a certain element of free riding or free lunch, vouchsafed by the existence of the nation-state, with its political and military might, which the AI startup types typically stay away from. Palantir, founded by Karp, Pieter Tiel, and others, has filled that vacant space, an intersection of the free enterprise and the harsh reality of the world. There remain enigma and scope for investigation in the present-day Leviathans that are the nation-states. Countries such as the United States, China, Russia, and others are constraints on human activities, and they would not disappear into the air no matter how your mindset might tell otherwise. Even in the free-for-all zeitgeist of TikTok and ChatGPT, you would ignore the Leviathans at your peril. On the other hand, there is something deeply counterintuitive about the nation-states. The spirit of Prometheus that is bringing a new fire to humanity does not thrive at the center of the Leviathan. There are certain truths about the idea that a nation-state would be the punisher, rather than the promotor, of the free flight and dance of Prometheus. In an ideal world, Prometheus would afford to be blissfully unaware of the pressure of the Leviathan. But of course, we are not living in an ideal world. We never did, and perhaps never will be. It is genuinely interesting to consider the makeup of decision making platforms that Palantir and other companies are developing, whether or not they are related to military/security purposes, the muscle and bone of the nation-states. There would be much scope for extended agency beyond that of the individual humans, and Karp and his colleagues are in an interesting ballpark. There is a clear undertone of The Technological Republic that leaves a strong aftertaste, suggesting the ultimate preferability of the Western ideology of democracy and free market, as an extension of the Francis Fukuyama vision at the end of the Cold War. Perhaps it is true that we have lost the polar star in the midst of academic fads and waves of political correctness. On the other hand, technologies such as AI do seem to tango with the authoritarian governments as well. Whether the technological republic to come will turn out to be benign or malicious is still an open historical question. Issues raised and hinted in this treatise would provide foods for thoughts for those concerned, in all walks of life, beyond that of the technophiles. The Technological Republic an important book that provokes new thoughts rather than settling to fixed ideas, which is always great in a time of transitions such as this. Reviewed by Ken Mogi @kenmogi

@JoelPenaMunozJr Yes I agree. Cognition is emergent phenomena.

A system survives by regulating itself between coherence and entropy, stability and novelty, order and uncertainty. Modern work in information theory, biology, and complex systems increasingly points toward the same idea: the “individual” may be a temporary equilibrium of information held together through feedback and interaction. Cognition may not exist outside physics. It may be one of its emergent organizational laws. I’ve called this the Equilibrium Field of Cognition. Really fascinating watching modern science slowly move toward information, structure, feedback, and dynamic equilibrium as deeper explanations for intelligence, identity, and reality itself.

@JoelPenaMunozJr Thank you for the message. You are right. How software is self-organized in biological systems is what I want to see.