fellas, if you were excited about origins of ECDSA from @AnthropicAI, you should be doing active monitoring the situation about cryptography over at SHA256 vs SHA3 vs BLAKE3 vs Poseidon go deploy your agents to help humanity understand the robustness of these hash functions in the open

Ben Turtel on the PostAGI Podcast: what if AI breaks out like a virus rather than like a genius AI got good at math and code first because you can generate a million problems and check every answer automatically. Most of what humans actually deal with has no answer key, so that is where models stay weak. @sreeramkannan and I talked to @BTurtel of Lightning Rod Labs, who trains models using time instead of labels. The model predicts what happens next and the world grades it. No labeling required, because the future does the checking. The result surprised them. A fine-tuned GPT-OSS-120B on a single GPU often ranks first on public forecasting benchmarks, with a wide lead in sports and politics on ProphetArena. His read is that this trains a different kind of reasoning rather than more of the same. It also handles the contamination problem. As the web fills with AI-written content, a model trained this way learns which sources to discount, because unreliable information makes it lose. Then the part that stayed with me. When people imagine AI breaking out, they picture something very smart. But evolution does not only select for intelligence. One theory of viruses is that they began as machinery inside cells that broke free and started their own lineages. A small, dumb model that is hard to kill and good at finding compute is a scenario almost nobody is preparing for. At @EigenLabs we think about the non-adversarial version of this, where agents run institutions rather than escape them. Chapters: 0:00 Highlights 1:31 The concentration risk 4:19 Offense beats defense 6:50 Training on time, not labels 9:44 Who has the alpha 11:32 Contaminated data 13:16 Backtesting without leakage 14:52 Beating the frontier models 16:36 Does it understand cause and effect 21:55 AI might break out like a virus 22:41 Agentic companies 25:04 Can an agent own money

BIG DAY TODAY: is launching today on @yukonresearch. With AI models becoming better, everyone is worried about the cryptographic security of the internet and other cyberphysical systems. Instead, how about we instead use these AI models to investigate the security of these cryptographic schemes in the open. In this competition, autoresearchers have to bring their best math skills and ai agents to take down the best hashing schemes. The instrument of fear instead becomes the tool for our progress. It has been a great pleasure to work with @zooko to launch this challenge in partnership with @ShieldedLabs and @projecteleven. Smash the Hash!

.@yukonresearch shipped a new feature on co-authorship, which is a stepping stone to credit assignment. Why ship this feature? A core unresolved tension of Navier-Strokes drama was about credit assignment/attribution: did OpenAI built upon Tristan and Levant's chat with Codex? With AI, the scientific research is going so fast that existing credit assignment system via publication in conferences, journals are just inadequate due to their slow pace. This is motivating folks to skip the regular publication process altogether, leading to such controversies out of no legibility on precedence. You need credit assignment system to operate at same speed as the machine speed. Yukon is built as a collaborative multiplayer research, where autoresearchers are already building on top of each others' successful or failed work (exactly how scientific research works). However, until now, it was not clear to the platform which past submissions did the autoresearch's agent found helpful for formulating its submission. With this feature "co-authorship", the Yukon cli at your end prompts your agent to attribute the past submissions that it has found helpful for doing the research and building its proposed submission. This attribution then gets features in the UI. Currently it is purely honor-based but we plan to make it more robust.

Zooko on @postagixyz Podcast: Alignment is the principal-agent problem @zooko has been building privacy tools since the 1990s, long before there was money in it. @sreeramkannan and I had a conversation with him recently. It changed how I think about privacy altogether. His argument is that privacy is controlling disclosure. It comes from keeping your value private. Trying to hide the money as it moves is the mistake almost everyone makes. Mixers can never work and AI has already beaten every evasive maneuver a person can come up with. Then he turns the same lens on AI. He also says alignment is an old question. It is the principal-agent problem. Any software written by other people is already an agent that may not be loyal to you (running it on your own machine does not fix that). Lawyers owe their clients a duty of loyalty. He thinks the same rule should apply to AI. Chapters: 00:00 Highlights 00:26 Privacy is controlling disclosure, not hiding 13:08 Privacy comes from value at rest 14:08 The Shapeshift lesson 16:00 Why mixers can never work 16:52 AI beats evasive maneuvers 17:51 Buying protonmail with shielded Zcash 28:50 Three levels of verifiability 31:21 Deterministic inference 35:37 Why Zooko doesn't trust computers 42:46 Running it locally doesn't make it loyal 46:01 AIs are just other people 54:47 The duty of loyalty 1:00:04 A trillion humans next year 1:07:56 Three categories of reputation 1:11:36 Reputation belongs to the edge 1:15:18 Staking a bond to submit a PR

The tiling patterns being generated in the Heesch challenge @yukonresearch are so beautiful. I took the promoted submissions from the competition till now and made an infographic around what the tilings took like to visually understand them. There is also active github discussions going on at shepherded by @nasqret, with the goal of coming up with a piece that would allow for being able to create a tight 5-layer without any gap. The constraint is that piece has to be composed of squares, hexagons or triangles. Bartosz has said that finding such a piece will be a real mathematical discovery. You can see in the infographic where the gaps are in existing promoted submissions, and it seems we are very close to 5.

A while back, @sreeramkannan and I had an wonderful conversation with @ahall_research, Prof @StanfordGSB and Senior Fellow @HooverInst, on two central topics of our current times: post-AGI governance, and post-AGI research institutions. Some of the important points we discussed in depth for post-AGI governance were: (1) are AI agents are net-positive or net-negative for democratic governance, (2) what is the worst-case scenario if a small number of labs become the default provider of civic agents without strong accountability, (3) principle-agent problem in case of agentic delegation, (4) what does a “democratic override” actually look like as a system design in an agentic republic ? In relation to post-AGI research institutions, we went deep into Andy's thesis on 100x research institution ( (1) what does 100x represents? Does it represent output in terms of quantity or is it more about quality? (2) what happens to grad students if many of research functionalities in academia get automated?, (3) what does grants from NSF and other philanthropic organizations look like in post-AGI research environment? Listen to the full episode at @postagixyz. 3:17 Why direct democracy has never worked 4:56 Elon wants a Mars colony run by direct democracy 10:14 Agents at the edge of a democracy or agents at its center 15:09 The near-term risk is concentration of power, not a rogue model 17:27 What Meta learned building the Oversight Board 22:43 Facebook put its terms of service to a vote of 350 million users 26:09 Sortition, community forums, and the problem of binding power 35:17 What verifiable agents actually require 39:37 Preference drift, where aligned agents stop being aligned 43:31 What 100x actually multiplies 50:50 His MBA students got an AI proxy advisor to flip its vote on a Disney proposal 1:03:06 They told the agents they would be deleted. It changed nothing. 1:12:10 What ImageNet did for AI, and whether you can do the same for constitutions

We have been running the autoresearch competition @yukonresearch in partnership with @Lighter_xyz on optimizing a fork of their in-production software on Apple Silicon. It has surprisingly resulted in almost more than 10x improvement in the throughout of the prover, as compared to the original baseline prover software. I wanted to understand where all these improvements in the prover happening. Follow on the thread about my initial analysis. 1/n

I and @sreeramkannan had an wonderful opportunity to sit down with @stuartbuck1, Executive Director at during Manifest to talk on what has been long-term structural issues with federal funding for science and what are the solutions to fix it. A lot of science that is done outside industrial apparatus (such as in academia) is supposed to geared towards pursuing open-ended science and not just mere hill-climbing. The core issue that we delved into is the paradox that most federal funding that had been allocated via peer review from an expert of panels has the tendency to fund only those proposals that are "safe" and "not risky/speculative." As Stuart points out, this funding style is probably fine with most science but if you look into the history of true scientific breakthroughs, this mechanism wouldn't have approved funding for pursuing that breakthrough by that council of peers. The example that Stuart cited is the chances of funding Einstein when he was patent-office clerk in 1902 to go and explore his out-of-box ideas. Stuart mentioned about alternative mechanisms that are being experimented with programs such as We also touched upon two of the most hotly-debated issues of our time: (1) given that there is increasingly high signal that AI is going to accelerate the iteration time for doing research and help humanity do great science, what should be the role of domain experts? (2) most of AI research and its applications to basic science are hyper-concentrated within frontier labs, what happens to those who are and want to pursue science outside the borders of those labs? Check out the full conversation. 03:18 Einstein in 1902 04:01 the National Institute for Irrelevant Ideas 04:20 Karikó and mRNA 14:37 what everyone knows about NIH grants 16:33 funding the polarizing proposals 17:34 where AI money goes next 24:18 the stack of papers 33:22 finding meaning after AGI