Twitxy Saved

Arun Rao

Verified account This account is verified. Learn more

@sudoraohacker

Builder of large-scale ML systems; adjunct prof @ucla; ex quant derivatives trader & startup founder. Tweets on AI, tech, science, & econ.
Berkeley, CA raohacker.com Joined June 2012
3.7K Following 5.7K Followers 18K Posts

Latest posts 5

Pinned post

Arun Rao Verified account @sudoraohacker
My favorite books and reading trace for 2025:

I had a five-way tie for the best books this year:

1) Mokyr’s A Culture of Growth on how societies build or block the conditions for sustained progress;

2) John Williams’ masterpiece Augustus, and the runner-up, Stoner, as intimate studies of character under the slow pressure of duty and power;

3) Dhami and Sunstein’s Bounded Rationality as the clearest bridge from “humans are cognitively weird” to “policy should be designed accordingly”;

4) Liu et al.’s Advances and Challenges in Foundation Agents as a big-map survey of how modern agents stitch together memory, planning, tools, and safety constraints into something that actually acts in the world, and;

5) von Neumann’s core computing writings (Compendium, EDVAC, Self-reproducing Automata, and The Computer and the Brain) arguing that computation is best understood as a class of automata with robustness coming from adaptation and distributed memory.

Links below.
3 replies 4 reposts 35 likes 15K views View on X
Arun Rao Verified account @sudoraohacker
This claim about the GRH will take 3-5 years to verify and maybe longer to understand-so sadly the claim isn’t as big as we’d like?
Mehmet Mars Seven 🐴 Verified account @MehmetMars7 4h
🚨 JUST IN: Turkish-American mathematician Levent Alpöge, working with Claude, claims to have proved one of the hardest problems in mathematics: the Generalized Riemann Hypothesis!

The Riemann Hypothesis is one of the seven Millennium Prize Problems.

The announcement includes a massive Lean formalization, which, when written out in English, runs to over 250,000 pages! 🤯

If verified, this could be the most significant mathematical breakthrough achieved with AI to date.
View quoted post on X
Arun Rao Verified account @sudoraohacker
I’m a fan of the Newsom/Becerra AI plan as I think it would encourage more labs and AI employees to exit California, which I think is healthy because the state needs to have less concentration in AI (and also needs less AI contribution to its bloated tax revenue, and fewer skilled AI workers).

More seriously, when writing regulation they need to consult the technical experts and business people working on this stuff - if they only let external lawyers yolo regulations on the book, just like they did with housing, you’re going to get bad outcomes but much faster and more drastic (industry exit).

But at this point I’m advocating for the AI industry to leave California and diversify, so perhaps getting some draconian laws enacted would be a good thing? (The federal proposals are much saner btw - it’s also why we need full preemption).
Jeremy B. White @JeremyBWhite 1d
NEW: Gov-elect @XavierBecerra floated criminal liability for rogue AI agents, telegraphing his plans for a top issue.

“Because it’s a machine doesn’t mean you shouldn’t also have the ability to prosecute when malfeasance, when bad things are done.”

politico.com/news/2026/11/1…
View quoted post on X
Arun Rao Verified account @sudoraohacker
Congratulations to Anthropic and those that built this lab - I’m proud of you all.

You built one of the great businesses in the history of capitalism and I know most of you don’t care and are trying for the bigger moonshot of ASI that shifts humanity to another level beyond petty squabbles of tribes, nation states, or economic scarcity.

I didn’t sign the petition against you as I know what the pressures are like in running a lab and making fast decisions with extreme uncertainty. I hope you publish more to bring others along with you, take the criticism thoughtfully as a member of Big Tech (where being criticized daily is part of the game), and play a more active leadership role in releasing your methods, data, and algorithms for safety and positive alignment so everyone benefits in a cooperative, alignment race to the top.

Peace be with you!