Phil Beaudoin
banner
beaudoin.social
Phil Beaudoin
@beaudoin.social
Researcher, coder, entrepreneur, kind. CS PhD, ex-Google, ElementAI co-founder. En français: @fr.beaudoin.social Current project: https://numeno.ai #FreeOurFeeds
Pinned
I wanted to spend the summer meeting inspiring people. Only something exceptional could have stopped me. Yoshua Bengio’s team at LawZero is this exceptional thing. I’m excited to be joining you all!
We are thrilled to welcome Philippe Beaudoin to LawZero as Senior Director, Research.

A seasoned researcher and entrepreneur, his experience will be invaluable as we advance our mission to build safe-by-design AI systems.

Full press release: lawzero.org/en/news/lawz...

@fr.beaudoin.social
Reposted by Phil Beaudoin
Notre directeur principal, Recherche, @beaudoin.social, est passé au balado Mon Carnet de @guglielminetti.bsky.social pour discuter de LoiZéro, de notre vision pour une IA sécuritaire, et du parcours qui l’a mené à ce nouveau chapitre.

Épisode complet : soundcloud.com/moncarnet/mo...
Mon Carnet du 1er aout 2025
Mon Carnet, le podcast de Bruno Guglielminetti Le grand magazine francophone de l'actualité numérique Vendredi 1er août 2025 Débrief transatlantique avec Jérôme Colombain (3:32) Entrevues : Pour un
soundcloud.com
August 1, 2025 at 7:29 PM
I’m working on a splaining Linux command just so “man splaining” can give you pages of information in the most annoying voice.
July 18, 2025 at 12:03 PM
The original excuse for my trip to London was to meet my friend @sherifea.bsky.social — we had not seen each other in five years!

« There’s this person I’d like you to meet, what about we record a podcast? », he said.

That’s how I met Vib. A collision I will remember for a long time.

(Links next)
July 17, 2025 at 11:08 AM
I wanted to spend the summer meeting inspiring people. Only something exceptional could have stopped me. Yoshua Bengio’s team at LawZero is this exceptional thing. I’m excited to be joining you all!
We are thrilled to welcome Philippe Beaudoin to LawZero as Senior Director, Research.

A seasoned researcher and entrepreneur, his experience will be invaluable as we advance our mission to build safe-by-design AI systems.

Full press release: lawzero.org/en/news/lawz...

@fr.beaudoin.social
July 16, 2025 at 1:35 PM
Seek out the strange conversations.

I just spent more than one hour discussing with Bishop Thomas Dowd.

1/🧵
June 8, 2025 at 9:19 PM
I’m don’t often paste inspirational quotes on my feed but when I do…
June 5, 2025 at 4:00 PM
Imagine a researcher with a desk just outside the cosmos, peering through a microscope at everything below.
On that desk lies a slim textbook titled:

Signs of Conscious Life.
June 1, 2025 at 2:57 PM
Pantone color of the year: Orange TACO.
May 31, 2025 at 3:24 PM
I fell into deep dialogue with language models.
I didn’t plan to, but I stayed.
And this is what I brought back.
fnos.ai
Field Notes on Something — A Different Kind of Inquiry on AI Systems
A personal and philosophical exploration of long-form dialogue with AI systems — written from lived experience, not theory.
fnos.ai
May 16, 2025 at 12:56 PM
Will we know when we've created something conscious?
In 2018 we went to Japan. Eve-Marie, the kids and I. It's one of our most memorable family vacations. It was filled with adventure, discovery, drama... love. 🧵
March 27, 2025 at 2:58 PM
Happy to spend my weekly meme budget on this one.
March 26, 2025 at 2:38 PM
Crappy code makes the world go round.
It’s how most SaaS companies get started, what powers many indie games, and it’s the first line you write before tumbling deeper down the rabbit hole.

Crappy code is often the fastest way to reach your goal — and it definitely has a role to play. 🧵/4
March 23, 2025 at 1:14 PM
A well executed, purely informative video. I doubt it will change minds but it will improve your ability to talk about the current reality in a clear way.
Love Wired's Tech Support series! This one is particularly relevant, given the recent actions of the current US administration.

www.youtube.com/watch?v=vK6f...
History Professor Answers Dictator Questions | Tech Support | WIRED
YouTube video by WIRED
www.youtube.com
March 23, 2025 at 1:16 AM
“I’m vibe coding!”
March 22, 2025 at 9:27 PM
“AGI preppers” is the best name I’ve read for this group. The problem with preppers is that they are so convinced of the magnitude and nearness of a change that they’re ready to face it alone or in a small group. Democracy is not a priority.
Should AGI-preppers embrace DOGE?
There's magical thinking about the magical thinking
www.programmablemutter.com
March 20, 2025 at 11:51 AM
Reposted by Phil Beaudoin
This is completely fucked up. I know everything is fucked up, but THIS is why universities need to stand up and fight, not cave like Columbia did. NOTHING will satisfy them, and all the capitulation just leads to more demands and humiliation.
I’m told this is a real letter. It basically says, “We’ll destroy Columbia unless you destroy it first.”
March 14, 2025 at 12:42 AM
Reposted by Phil Beaudoin
BREAKING: Hundreds of Jews and allies have taken over Trump Tower chanting “We want justice, you say how. Bring Mahmoud home now!” and “Fight Nazis, not students.”

The civil disobedience is spawned by the ICE arrest of Palestinian student Mahmoud Khalil.
March 13, 2025 at 3:51 PM
Few things are more concerning that people in power lacking empathy. One such thing is people in power actively fighting empathy in others.
March 9, 2025 at 10:39 PM
Wild thought: The best way to protect democracy might have been… to elect Trump in 2020.
March 6, 2025 at 1:31 PM
Doubt the Founding Fathers had 'USA-Russia-China axis' on their 250-year bingo card.
March 5, 2025 at 5:58 PM
It’s happening fast. “Reasonable people” have trained themselves to identify a feeling of rapid change as “an internal emotional response”. As a result we imagine we are probably over-reacting…

…maybe not. These are not normal times.
March 5, 2025 at 12:16 PM
The idea of doing science without having an interest in philosophy is weird to me. It’s like having the best tool in the world but not caring how it was made.
February 27, 2025 at 1:41 PM
Two AI risks in tension with one another: 1) existential risk (loss of control, malevolent use); 2) risk of power concentration.

Who needs any more proof that power concentration is the most pressing one?
February 26, 2025 at 7:01 PM
Days since I typed a ChatGPT query into a chat with friends: 0
February 21, 2025 at 3:23 PM
I call this PR "The Ship of Theseus".
February 21, 2025 at 1:04 AM