Ethan Siegel has a discussion of “vibe physics“, people convincing themselves that they can solve fundamental scientific problems by chatting with an LLM. For a story about one billionaire doing this, see here. LLMs should be much better than the usual crackpots at generating worthless papers about theoretical physics, likely should be able to generate papers not easy to distinguish from a lot of what is on the arXiv. I’m wondering how much of this has already happened.

In math, Daniel Litt has noticed a bunch of recent LLM-generated worthless papers on the Hodge conjecture. As examples, he points to these, four papers posted during the past month. Unfortunately the arXiv does not seem to now have an effective way to protect itself against these things getting posted, or to get them removed once identified (Daniel identified them publicly two weeks ago, no indication anything will be done about this).

All of which brings us back to the notion of vibe physics. Sure, an LLM may sound very intelligent and knowledgeable about theoretical physics, particularly if you yourself aren’t an expert in the areas of theoretical physics that you’re discussing with it. But this is exactly the problem any non-expert has when talking with:

• a bona fide, honest, scrupulous expert,
• a dishonest grifter posing as a scrupulous expert,
• or a confident chatbot that has no expertise, but sounds like it does.

The problem is, without the necessary expertise to evaluate what you’re being told on its merits, you’re likely to evaluate it based solely on vibes: in particular, on the vibes of how it makes you feel about the answers it gave you.

This is the most dangerous thing for anyone who’s vested in being told the truth about reality: the potential for replacing, in your own mind, an accurate picture of reality with an inaccurate but flattering hallucination. Rest assured, if you’re a non-expert who has an idea about theoretical physics, and you’ve been “developing” this idea with a large language model, you most certainly do not have a meritorious theory. In physics in particular, unless you’re actually performing the necessary quantitative calculations to see if the full suite of your predictions is congruent with reality, you haven’t even taken the first step toward formulating a new theory. While the notion of “vibe physics” may be alluring to many, especially for armchair physicists, all it truly does is foster and develop a new species of crackpot: one powered by AI slop.

The rise of AI is breeding a new generation of physics crackpots who get endless positive reinforcement from ever-willing large language models. This one takes it to a hilarious new level.

.....

Collaboration Request | Testable AI-Driven Theory of Everything

Hello, I hope this message finds you well.

I have formalized a meta-methamatical theoretical framework—Unified Quantum Relativity Theory (UQRT)—that unifies relativity and quantum mechanics into a coherent set of equations. With the assistance of AI, I have been able to derive equations that, while not yet fully rigid, hold significant promise for unifying these two foundational pillars of physics.

To explore this further, I invite you to test UQRT yourselves. Simply upload the attached manuscript to your AI systems and step-by-step instruct them to: “(1) read it section by section, building understanding progressively, (2) derive the general equations and the unified equation based on the framework progressively—not from the outset, as the paper reinterprets concepts in ways that diverge from conventional approaches until the full picture emerges, and (3) derive all laws of physics from the unified equation and compare these equations against experimental results from cosmology, blackhole physics, etc.”