Charlotte Peytour

head of marketing & ops @wetransfer @bendingspoons - philosophy & tech - founder of lophi, media dedicated to the future of humanity -

[Sigal Samuel @SigalSamuel]

I reported this deep dive for @voxdotcom because I think more folks really need to be aware of this — here’s a gift link:

https://t.co/gXd6g7Dnke

2/n

[Le Grand Continent @Grand_Continent]

Magnifica Humanitas

Plus de cent pages, 250 paragraphes, près de 40.000 mots.

Nous publions le texte intégral de la première encyclique consacrée à l'IA.

Le premier texte doctrinal du premier pape américain commenté ligne à ligne.

À lire absolument.
https://t.co/vajiUXqEhu

[Charlotte Peytour @Charlotte_Pey]

Enfin, pr terminer: le pari open-source chinois, les scientifiques artificiels - probablement ma partie préférée, à savoir le développement d'agents capables de traiter des questions de recherche complexes. Et, évidemment, les prémices d'une révolte politique contre l'IA.

[Charlotte Peytour @Charlotte_Pey]

10 sujets à suivre dans l'IA selon @techreview (article du MIT)

1️⃣ Les données humanoides - ou ces millions de vidéos humaines plus qu'étranges, où des petites mains exercent les gestes du quotidien, fournies pour entraîner les robots

https://t.co/M0kPEPyN2F

[Charlotte Peytour @Charlotte_Pey]

Sans oublier: l'omniprésence des outils d'intelligence artificielle dans la prise de décision militaire, le développement de deepfakes, l'orchestration de plusieurs agents en même temps pr atteindre des objectifs encore plus complexes.

[Charlotte Peytour @Charlotte_Pey]

👀

[Owain Evans @OwainEvans_UK]

As AI systems are increasingly trained on one another’s outputs, subliminal learning means they may inherit properties not visible in the data.

[Charlotte Peytour @Charlotte_Pey]

Et bcq d'autres thèmes sont abordés - ex de Severance et persistance de l'identité personnelle, question des masques et du Shoggoth (un de mes passages préf)... J'en parlerai dans ma newsletter https://t.co/S06Vix3pqq dont je prépare les prochains numéros 📝

[Charlotte Peytour @Charlotte_Pey]

David Chalmers vient de publier "What We Talk to When We Talk to Language Models." Un papier dense mais brillant qui pose la question suivante : quand on parle à Claude ou ChatGPT - à qui parle-t-on vraiment? Et "tue-t-on" un modèle de Claude quand on ferme une conversation?

[David Chalmers @davidchalmers42]

here's a new version of "what we talk to when we talk to language models", with an added section (pp. 16-23) on LLM interlocutors as characters, personas, or simulacra. https://t.co/RLDP5FFmgM

the new version discusses role-playing vs realization, the simulators framework, the persona selection hypothesis, and more -- in addition to the existing discussion of quasi-mental states, LLM identity, personal identity in severance, LLM welfare, and related topics.

this version was mostly written before recent discussions of these issues on X and in NYC, but i've updated it a little in light of those discussions. any thoughts are welcome.

[Charlotte Peytour @Charlotte_Pey]

Cet Assistant est "sticky". Mais du coup, est-ce encore un rôle, ou est-ce que le modèle est devenu ce personnage ? C'est la différence entre un acteur classique qui joue Hamlet et un method actor qui devient Hamlet en vivant toutes ses émotions, nous dit Chalmers.

[Andrej Karpathy @karpathy]

Judging by my tl there is a growing gap in understanding of AI capability.

The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code.

But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along.

So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions.

TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.

[Serena Yan @ysl_serena]

Physical Intelligence is the next OpenAI. Mark my words.

[Tim Urban @waitbutwhy]

Good day to remember just how big that green tree is

[َ @o_exacoustos]

Cette phrase que Friedrich Nietzsche avait écrit à son ami Jakob Burckhardt : « J’ai atteint le point où je vis comme je pense. »

[Infinite Books @infinitebooks]

Easy

[Stuz0r @Stuz0r]

A collection of my Unreal Engine works over the past few weeks.