George Adamopoulos
47 posts
George Adamopoulos
@GeorgeAdams101
Person who happens to be living through the singularity
Katılım Kasım 2022
925 Takip Edilen242 Takipçiler
@RichardMCNgo The open endedness research (Why greatness cannot be planned) is a convincing argument that theory of impact is generally a bad thing
English
One reason I hate the idea of having a “theory of impact” for your research is that it limits you to research with predictable consequences.
But that rules out research which aims to improve our understanding of foundational concepts, which is the most important kind!
English
@savinduwim @AndrewCurran_ Just as nick land predicted
English
@AndrewCurran_ AI taking jobs before job automation to fund data center capex
English
META is about to make large scale layoffs to offset AI capex that could affect more than 20% of the company.
English
@LaurenceLiang1 @mcgillu depending on when it is, i might be interested in helping out
English
a good friend of mine is considering throwing an AI research hackathon at @mcgillu this semester, if there's a strong enough demand
is anyone interested in helping co-host? happy to connect!
English
This was fun. Thanks for coming everybody!
steve hsu@hsu_steve
Nick Land in SF: Acceleration FTW! @xenocosmography
English
@brianryhuang @taifuunoyume In a sense, if I take your premise of no AGI, the value of a frontier lab would be data and having made model improvements some number of months before open source. And also distribution.
English
@brianryhuang @taifuunoyume I buy the premise that Opus 4.5 is ~transformational for SWE. Why does Anthropic necessarily have a moat that another company can't take? How long until there's an open-source version?
English
i'm pretty risk-averse with investments and have never touched crypto, but i'm just saying i'm 3x levering on my life savings into the anthropic IPO next year
English
@taifuunoyume @brianryhuang @GeorgeAdams101 The possibility of AGI shouldn’t really interact with your bullishness on a particular AI lab stock
English
@peterjliu What does this say about human level intelligence :)
English
It's better to rent than to own a rapidly depreciating asset
Hieu Pham@hyhieu226
Uber has no cars. Airbnb has no real estate. What AI agent company owns an LLM?
English
@brianryhuang Is P(anthropic builds AGI and you 50x your return) > P(AI bubble)?
English
George Adamopoulos retweetledi
One must imagine Sisyphus happily reposting this
Financial Physics@FinancialPhys
If capitalism lifts people out of poverty Why is there so much poverty?
English
English
There are no words to describe what I am feeling right now, what a huge honor.
Thank you everyone at @arcprize, especially @fchollet, @GregKamradt, @mikeknoop for choosing TRM for the 2025 #ARC Paper Award!
Special thanks to @k_schuerholt for replicating TRM! 🙏
ARC Prize@arcprize
ARC Prize 2025 Paper Award Winners 1st / "Less is More: Recursive Reasoning with Tiny Networks" (TRM) / A. Jolicoeur-Martineau / $50k 2nd / "Self-Improving Language Models for Evolutionary Program Synthesis: A Case Study on ARC-AGI" (SOAR) / J. Pourcel et al. / $20k 3rd / "ARC-AGI Without Pretraining" / I. Liao et al. / $5k
English
@IvanVendrov Your essays are awesome. I have high expectations lol
English
@MechanizeWork Why can't you have a sweet of open source RL envs like a "pretraining" corpus? and then also have specialized envs. It seems to me the benefits of open source software hold up for RL envs as well
English
Open source is not a viable strategy for producing RL environments at scale. The future of AGI lies in closed source development.
Matthew Barnett@MatthewJBar
The AI industry seems biased towards open source development, even though it keeps failing to deliver. OpenAI was founded on the idea of open source, only to abandon it. Meta backed open source, but now seems to be walking back. Mistral has barely had any market impact.
English
So excited to join the new @SimonsFdn Simons Collaboration on the Physics of Learning and Neural Computation, to further advance our understanding (mech interp) of learning & reasoning in large networks, including classical deep nets and other bio-inspired network models!
Simons Foundation@SimonsFdn
Our new Simons Collaboration on the Physics of Learning and Neural Computation will employ and develop powerful tools from #physics, #math, computer science and theoretical #neuroscience to understand how large neural networks learn, compute, scale, reason and imagine: simonsfoundation.org/2025/08/18/sim…
English
George Adamopoulos retweetledi
I've been working on a @3blue1brown style video explaining our research on attention patterns of OpenAI's Minecraft agent VPT.
1min teaser below.
You can see a longer WIP version in my recent talk at @cohere (from 24:55):
youtube.com/watch?v=p_vwKf…
YouTube
English
George Adamopoulos retweetledi
Announcing the AI Futures Project: an AGI forecasting and governance organization led by @DKokotajlo. Our first project is AI 2027: a scenario forecast of the development and effects of superintelligence. We've also developed an AGI tabletop exercise.
English