gave a short lecture this morning on principles of computational modelling, always try to stress the point made by @romainbrette.bsky.social that adding details to a model does not automatically make it more realistic. The wooden airplane model has more 'details' but only the paper model can fly

AI slop and the destruction of knowledge irisvanrooijcogsci.com/2025/08/12/a...

Why large language models are not conscious. romainbrette.fr/notes-on-con...

Two b, or not two b, that is the question.

I guess a PhD is all about questioning what everyone takes for granted.

At this point, I wonder why one would want to justify their neuro research by the hope of making a breakthrough in AI. Like, that would be a good thing?

so true

Banning for-profit journals won't solve everything, but it's a necessary first step. One thing in particular that publisher money is consistently used for is lobbying against the general interest, and this is one of the greatest obstacles to political change.

Such a great paper, blowing the false dichotomy of genes vs. environment. Primate brain folds, connectome and even behavior is better predicted by brain volume than genes. Folds appear as a result of development with mechanical constraints (big brains get folded). No map of the brain in the genome!

I believe at this time we can afford more nuance and precision than this. Most null results are uninformative because most experiments are uninformative because we use NHST to replace the scientific process. We can't keep avoiding the root problem forever. Also the focus on efficiency is jarring.

I changed my mind. The tech bros might actually be on their way to achieve superhuman artificial intelligence, but by making people dumber! What a brilliant strategy!

Dishwashers are pretty good at washing dishes! I wonder how the human body implements water-spraying rotating arms.

Against "reverse-engineering the brain": (The Brain, in Theory, chapter 1)

Why a mental state is not some kind of physical state. (The Brain, in Theory, chapter 8)

Open access edition of "Principles of Biological Autonomy" can be found here: direct.mit.edu/books/oa-mon...

Theoretical cellular neuroscience please! Theory of development very much needed too! One question I find particularly interesting: how does a complex "instinct" develop, say the spider's web making?

We are super happy to announce the third Workshop of Ideas in Neuroscience! We will once again look critically at assumptions of modern neuroscience: what does it mean that the brain encodes information? Is this a useful approach, or a metaphor that blurs our vision?

Information is the epistemic phlogiston of neuroscience. (Chapter 6 of my book, "The Brain, in Theory")

What is computation? ("The Brain, in Theory", chapter 4)

Today is publication day! A new edition of Francisco J. Varela's (long hard-to-obtain) 1979 classic Principles of Biological Autonomy in a new annotated edition by Ezequiel Di Paolo and Evan Thompson, with a Foreword by Amy Cohen Varela. mitpress.mit.edu/978026255140...

New paper: Single camera estimation of microswimmer depth with a convolutional network We use a simple convolutional network to estimate the 3D position of a protist (Paramecium) from a single conventional microscopy image.

Because we must build good things while we scream about the bad, I have started a "Data for Good" team @data-for-good-team.bsky.social that partners with organizations needing short-term data science help. We have three projects ongoing & will add more as our capacity grows. data-for-good-team.org

Reminder that AI doesn't have "math skills" or a "handle on facts". Those are things that people have, not algorithms.

Principles of Biological Autonomy by Francisco Varela, new edition Now that biology is finally catching up with Varela (eg agency is commonplace and teleology not scary) the book is probably more timely than ever. Here's my physical version! The commentary by Di Paolo and E. Thompson very helpful.

This « Choose Europe for Science » event organized by Macron is an obvious cover up. Not *a single French academic* is attending, as it follows a series of unprecedented cuts in academic funding (-3.5 billions). www.education.gouv.fr/lancement-de...

Just published: "Theory of axo-axonic inhibition". With a little more content. journals.plos.org/ploscompbiol...

Why science is not reductionist, and never has been. (my forthcoming book "The brain in theory", chapter 3)

Un billet sur les attaques contre la liberté académique, en lien avec les réformes en cours en France. romainbrette.fr/science-et-d...

Suppression votée en séance Merci, bien joué @arnaudsaint-martin.bsky.social @hendrikdavi.bsky.social Les socialistes viennent de s'aliéner définitivement la communauté universitaire. Mais ce n'est pas très grave ! La lutte continue ! Navette vers le Sénat maintenant

Victoire !!! Nous venons d’empêcher le rétablissement du #HCERES ! On l’a supprimé ! #DirectAN

A good book! (TDLR: it doesn't)

BREAKING NEWS. Le PS retire son amendement, car «  Nous avons aussi conscience du fait que cette institution a été utilisée comme un moyen de contrôle budgétaire par le Gouvernement ces dernières années. Cela ne correspond pas à la vision que nous nous en faisons. »

La pétition pour la suppression du Hcéres a dépassé les 4000 signatures. Signez, faites signer si ce n'est déjà fait. Il reste moins de 24 h pour convaincre, pour que les députés PS retirent leur amendement et ne votent pas ceux du gouvernement demain 8 avril. Go ! rogueesr.fr/supprimons-l...

Cette interview est proprement hallucinante. Mme Chevallier succède à M. Coulhon, conseiller ESR d'Emmanuel Macron. Mais le Hcéres, c'est important pour la démocratie parce que c'est une institution "indépendante du pouvoir exécutif". Mais bien sûr.

Wow. On se demande dans quelle bulle vivent les socialistes @socialistesan.bsky.social. "Il y aurait donc un risque de retour à la complexité antérieure à la création du Hcéres." Euh, on parle bien de la surcouche bureaucratique délirante qui évalue les chercheurs à coup de diagrammes SWOT?

One thing I don't get re: Columbia. The university has ~$20B in assets, an increase >$1B in just a year. How's a $400m cut such a threat? Even if it were to last for the next 4 years, it could be absorbed by selling some stocks. www.finance.columbia.edu/news/financi...

What is the point of a scientific society that won't stand up against an attack on science by an incipient fascist? www.theguardian.com/technology/2...

Le rapport de la @ccomptes.fr sur le @cnrs.fr est pathétique: un constat souvent pertinent, des solutions inadaptées. Exemple avec les missions. La cour a raison, c'est une catastrophe. Mais les préconisations sont stupides. Non le CNRS et les universités ne peuvent faire mieux !

Les libertés académiques ne sont pas là pour assurer compétitivité, croissance ou souveraineté. Elles sont là pour assurer l’indépendance de la recherche et de l’enseignement supérieur vis à vis des pouvoirs. Et ce sont elles qui assurent un usage démocratiquement décidé des savoirs. #VeilleESR

Assez avec cette institution parasite de la science qui bénéficie d'un plafond d'emplois de 128 postes fonction publique et de 24 Millions de budget dont 14 pour les salaires (soit quand même plus de 100.000€ par poste) ! Le HCERES ne sert strictement à rien, donnons ses moyens aux universités.

Alors que l'on redécouvre que la science libre est un bien commun précieux pour la démocratie, et non simplement un incubateur de start-ups, je reposte ce texte de politique des sciences où j'explique que le fondement de l'activité scientifique est une éthique de la vérité.

Absolument. Pour le stockage, peut-être du peer-to-peer entre labos?

I'm giving an online talk today at 4 pm (Paris time): "Brains beyond computers". univ-cotedazur.eu/efelia-cote-...

Just wondering. Are preprint servers potentially at risk of being taken down or censored?

My Christmas gift to you fellow neurophysiologists: "Theory of axo-axonic inhibition". www.biorxiv.org/content/10.1...

Attention my fellow spiking neural network python nerds, there is a new version of Brian (2.8) just in time for Xmas. Particularly recommended if you use random numbers in standalone mode, big speed improvements. brian2.readthedocs.io/en/2.8.0/int... Brian will make his way to Bluesky soon. 🤖🧠🧪

I wrote a book: "The brain, in theory". First chapter and TOC: romainbrette.fr/WordPress3/w...