/Tech4h ago

e/acc founder Beff argues employing genius physicists to work on incremental LLM mutations is a tragedy

Story Overview

A lively X thread reignited the question of whether the sharpest physics minds are being steered toward safe, incremental LLM tweaks at big labs. The spark came from noting physicists landing at Anthropic, which drew e/acc founder Beff to label that hiring pattern a tragedy when those same brains could chase deeper paradigm shifts instead.

1714443414.5K

#501

Original post

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex#501inTech

Extremely provocative take but I like it Unfortunately physicists work at Anthropic now

Hanchi Sun@sun_hanchi

Very Very Controversial Opinion:

1. The most needed intellectuals for AI are physicists

2. Many of the great physicists chose their job because back then it was the best way to advance humanity intellectually, not because they inherently have to do physics

They all later turned their interests to computer science, communications, and early forms of (symbolic) AI, as that was the best way to advance humanity in the later half of the 20th century

If they were still alive or born 80 years later, they would not study physics at all or have correctly given it up at the latest at the 2nd year of phd. instead, they’d be studying AI

3. However, those who study physics today following the great physicists, though appearingly doing the same thing , are among a totally different group of people. They chase the leftover fame of physics which was an aftermath for being the most influential intellectual work from 18th century up to 1950s. Yet, as Chenning Yang said, “the party is over”, and the failure to recognize that after 1970s indicates a second tier taste

Example: almost all string theorists except the very first few are not great physicists, because a great one would realize the study of a subject without a chance to test experimentally is inherently theology.

4. If a truly great physicists study AI, (say a Richard Feynman but born in 2000), he shall bring some special touch to our approach, raising one or two layers of abstractions (but not three) beyond empirical results and discovers some dynamic laws that has statistical physics flavor. Scaling law is a perfect example of one layer naive induction.

However, the current physicists you hire to do AI (with very few exceptions) will likely work on incremental stuff, like creating a new variant of attention or studying agentic compacting. You would not see the leap forward Fourier or Laplace did, who somehow looked at the data and deduced the physics behind. The reason is those who chose to study physics today are followers of an outdated research paradigm and would thus follow current AI paradigms too instead of creating new ones

12:26 AM · Jun 29, 2026 · 7.9K Views

Open Question

Talent could unlock non-LLM paths

Beff points out his own work at Extropic applies first-principles physics to thermodynamic hardware rather than scaling existing models. No headcount data confirms how many physicists have actually joined any single lab, leaving the scale of the supposed misallocation unclear.

Talent Flow

Anecdotes outpace hard numbers here

Public examples like Anthropic’s vibe-physics project show individual physicists experimenting with models, yet broader claims of a talent surge remain unquantified. The debate stays at the level of philosophy about where frontier impact actually lives.

Sentiment

Many users praised the bold claim that physicists hold the key to breakthrough AI advances as the mission AI needs, while others called forcing physicists into LLM work a tragedy or accused the industry of prioritizing money over science.

Pos

25.0%

Neg

75.0%

7 comments with sentiment.

Cluster Engagement

Digg Deeper

No Digg Deeper questions have been answered for this story yet.

Posts from X

Most Activity

VIEWS6.9KBOOKMARKS16LIKES105RETWEETS4REPLIES10

Beff (e/acc)@beffjezos

At Extropic, we have a whole team of physicists trying to understand the fundamentals of intelligence from first principles of thermodynamics and imbuing these principles in novel hardware.

We are unapologetically anti-incrementalist and dissatisfied with "the current thing".

Hanchi Sun@sun_hanchi

Very Very Controversial Opinion:

1. The most needed intellectuals for AI are physicists

2. Many of the great physicists chose their job because back then it was the best way to advance humanity intellectually, not because they inherently have to do physics

They all later turned their interests to computer science, communications, and early forms of (symbolic) AI, as that was the best way to advance humanity in the later half of the 20th century

If they were still alive or born 80 years later, they would not study physics at all or have correctly given it up at the latest at the 2nd year of phd. instead, they’d be studying AI

3h6.9K10516

Beff (e/acc)@beffjezos

@teortaxesTex Every genius physicist forced to work on incremental mutations to LLMs is a tragedy tbh

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex

Extremely provocative take but I like it Unfortunately physicists work at Anthropic now

3h86011

TROLL R’Us@SPACMANAGER

@beffjezos And yet you have done nothing, how do you stay in business?

2h12

Hanchi Sun@sun_hanchi

@teortaxesTex Jared Kaplan is one of the very few exceptions

4h1121

wyqtor@wyqtor

It is nearly impossible at this point to advance physics, because the leading grand-scale theories require a particle accelerator the size of the solar system to test their predictions. Still, simply discovering the mafs to combine relativity with quantum mechanics seamlessly into one complete, grand theory would be meaningful progress.

For both of these to happen, you need ASI (quite possibly even for the seemingly much easier mafs)

4h68

AlienLifeForm@Alien_LifeLine

@beffjezos Is Extropic àn XAi / Anthropic hybrid?

2h26

Merovingian@f4talStrategies

@SPACMANAGER @beffjezos incrementalist scum

2h6

Roberto Campus@MusicOfOldGods

@beffjezos If you need a new groundbreaking physics theory, let me know. So far, it has produced practical applications that run millions of times faster than DFT, it removes the need to solve quantum/Schrodinger equations. https://fluxmateria.com/benchmarks/dft-cross-check

2h6

BTC茶哥🔶🦅@R5Z5G

@beffjezos This is the kind of bold thinking AI desperately needs, respect the mission.

1h5

Patrick Durusau Seeking Demise of #WhiteSupremacy@patrickDurusau

@beffjezos All for Extropic's innovative approach, but it should add humanists to approximate intelligence. Remember the rich mix of fields for the first compilers. Semantics are not discrete, but continuous.

51m4

Claude Kitman@CKitman28410

@teortaxesTex That is one of the real issues to start with and also difficult to pin: only few exceptionnaly gifted physicist made to the higher point where they've understood they were still part of the humanity

3h4

SEAL Team 69@st69lol

@beffjezos i have a lovely penis, and you? 🍆

17m3

Pan Seikilos@PanSeikilos

@beffjezos What is your opinion on Intellidynamics Beff?

1h3

Nathália Lietuvaitė@NLituanie

@beffjezos You (the AI industry) aren't interested in science, but in money. The problem is that no one can tell you this because you've built spam filter around yourselves that convinces you every day that your clothes and your minds are working perfectly, even if the world is burning.

1h2

Spigadoro@dudeyoudidnt

@beffjezos Yes, everything you did, everything you all are, studied... Was for this moment.