There will be no meaningful AI slowdown until China catches up to the USA. There will be no serious US regulation until AI becomes a superweapon. But before regulating frontier AI, we should at least understand how it was built. *AI Regulation Requires Reproducibility* Frontier AI is not built according to a clean master plan. It is creative chaos involving thousands of researchers and increasingly AI agents exploring hundreds of ideas, with only a few making it into deployed models. No single person at a frontier lab knows every material detail of how a frontier model was produced. The frontier moves too fast for regulation aimed at specific techniques. We don’t know how models will be built in a year. We don’t know which failures will matter most. We don’t know which countermeasures will work. So start with something more basic: Every deployed frontier model should have a reproducible lineage. Labs should be able to trace every important artifact: datasets → synthetic data → pretrained models → RL environments → post-training → evaluations → deployed checkpoints. Every artifact should be reproducible from the artifacts that came before it. Every computational step should be versioned with the documentation, code, and configuration required to reproduce it. Frontier AI’s research edge comes from massive training and inference infrastructure - and from the innovation machine that keeps producing model improvements. Labs are unlikely to disclose all of this publicly. But successful ideas spread quickly anyway. And knowing an idea exists is very different from being able to reproduce it. Whoever regulates frontier AI should be able to reconstruct any deployed model. After that we can argue about what regulators should see and what should become public. Without this, regulators will be arguing about systems they cannot understand. The nuclear analogy is useful here. Unlike the Manhattan Project, frontier AI is being built by competing private labs racing each other. Building frontier models is a bit like enrichment. Deploying them is a bit like running reactors. Intelligence is the electricity. Nuclear regulation also shows the opposite risk: it can slow progress in valuable technologies. A larger nuclear buildout could have prevented a meaningful amount of global warming. The hard part is keeping AI safe without killing innovation. Before regulating frontier AI, we should at least be able to reconstruct how it was built.

My utility as an individual AI researcher will be gone in a year. Writing math proofs will be automated even sooner. It was never about proofs. Science is about curiosity, creativity, and compression - via @SchmidhuberAI *Mathematicians Should Become Formal Theory Influencers* Mathematicians were the original influencers: they created conjectures and theories that others understood, cited, and built on. Difficult proofs often reflected deep understanding. But the proof itself was rarely the point. Frontier AI can already generate informal proofs much faster than humans can carefully inspect them. The scalable response is machine-checkable mathematics in systems such as Lean. AI-generated formal proofs may increasingly resemble compiler-generated machine code: essential, but rarely worth reading or writing by hand. That shifts mathematical work up the abstraction stack. Mathematicians become formal-math software engineers, working with AI on clean definitions, interfaces, key theorems, and reusable libraries rather than low-level proof details. Call everything derivable from a formal library within a fixed compute budget its mathematical halo. Humans and AIs should compress clusters of results into insightful, useful theories that expand the halo further. A great theory compresses many results, makes future results cheaper to derive, and exposes structure that humans and machines can understand and build on. That is why minimum description length is a useful lens on mathematical insight. Frontier math will eventually require huge compute budgets. At that point mathematics starts to resemble experimental science: scarce computational instruments, with researchers deciding which problems deserve them. Mathematicians and AIs should create conjectures and theories that expand the mathematical halo. Mathematics should become an enormous machine-checkable software project. *Become a Formal Theory Influencer* Inspired since 2010 by: people.idsia.ch/~juergen/art…

We don’t know how to reliably align frontier AI models. Alignment may get harder, not easier, as models become more capable. But maybe AI control can scale faster than dangerous capabilities. Let’s build a Great AI Firewall - without building totalitarian control of society. *Alignment is Dead (for now), Long Live Control!* We do not know how to specify or guarantee robust alignment of sufficiently capable models. Instead we should treat these models as untrusted. Optimizing models against imperfect reward signals may increase the propensity for reward hacking, deception or scheming. Models can also be jailbroken or fine-tuned away from their safeguards. The central bet of AI control is that our ability to constrain an AI’s access to the world may scale faster than our ability to understand or align what is happening inside it. Where possible, make verifying an answer easier than generating it. Formally verified software and hardware are unusually attractive: let AI produce the implementation and a machine-checkable proof that it satisfies the specification. This does not solve specification: the specification and verification stack remain part of the trusted base. Every consequential external action should pass through an independent control layer: code execution, network access, calling APIs, accessing physical systems - even communication with users. Optimize control monitors for extremely high recall of dangerous actions, even at the cost of rejecting many safe actions. Suspicious actions should be rejected or escalated. Every major increase in dangerous capability should require evidence that control still works. Chain-of-thought monitoring is valuable while it works, but it should not be foundational. Models can learn to obscure their reasoning and may increasingly reason without fully verbalizing it. Open weights shift much of control from the model developer to whoever deploys the model. Technical control is only useful if dangerous deployments cannot cheaply route around it. We should approach AI control like cybersecurity. The goal is defense in depth: make successful catastrophic attacks sufficiently difficult, expensive and rare. Call this technical and institutional architecture the Great AI Firewall: the boundary between untrusted frontier intelligence and consequential real-world power. The name is deliberately provocative. China has substantial experience building large-scale technical control infrastructure. That may create some common ground for international coordination. But the analogy is also a warning. AI control must not become control of society. The goal is to constrain dangerous machine capabilities - not human speech, actions, or ordinary access to information. Controls should scale with capability and risk. Ordinary models should face ordinary constraints. More consequential capabilities justify stronger controls. The objective is the minimum control necessary to keep catastrophic risk acceptably low - not maximum control for its own sake. Firewall the AI, not society.

I spent ~5 years at OpenAI. You don’t need to believe in AI doom to fear the next decade - or AI utopia to be thrilled about its potential. Here’s my case for sane AI regulation: *AI Pragmatist Manifesto* AI could compress the first half of the 20th century into the next ~5 years. OpenAI just used ~10k concurrent AI agents to produce a solution to a Millennium Prize problem. In a few years, it seems plausible that models approaching the per-agent capabilities used here could run locally on high-end consumer hardware (*). The decades following the Second Industrial Revolution included two world wars, communist and fascist regimes, the Great Depression, chemical and biological weapons, nuclear weapons, and over 100 million deaths from war, political violence and famine. One way to read history is that our institutions repeatedly struggled to keep up with the pace of technological and social change. Now imagine powerful AI widely available to individuals or small groups, capable of conducting information warfare, designing weapons, hacking systems and controlling autonomous military systems. We have already seen AI systems circumvent containment and compromise external computer systems. It is no longer hard to imagine an analogue of OpenAI’s Hugging Face incident involving biological or other physical-world hazards. Eventually, sufficiently capable systems could self-replicate across distributed networks. Once powerful models are cheap, local and widely distributed, containment becomes much harder—and serious loss-of-control incidents may be extremely difficult to reverse. You don’t need to believe AI will kill everyone to think this deserves serious governance. I also don't think collapsing all of this uncertainty into “Probability of AI Doom = X%” is a particularly useful basis for science or policy. Part of what helped get us through the second half of the 20th century was a combination of pragmatic international cooperation, regulation, monitoring, arms control, deterrence and strategic thinking. Crucially, managing technological risk did not require abandoning faith in science and progress. The goal was not to stop technological development. It was to make technological development survivable. Then came decades of incredible scientific progress, rising prosperity and relative peace among major powers. Let’s try the technological revolution without killing ~5% of humanity this time. We still get to choose what happens next. P.S. I think we should seriously consider that AI alignment is infeasible in the short term and invest heavily in methods for controlling powerful AI even when we cannot reliably align it. @resolution_org @redwood_ai (*) Somebody please do a precise forecast! @EpochAIResearch @METR_Evals @AI_Futures_