I interviewed @tsmilton_amodo about the tech we'd need to verify a potential US-China deal on AI. We discuss Amodo's prototype for inference verification, as well as the myriad ways you could use power monitoring, zero knowledge proofs and/or unilateral surveillance to understand how an adversary is using compute. Though solving some kind of verification is central to many people's theories of victory, there are precious few people working on it currently. Tom is one such person! Listen to the episode on Youtube, Spotify or Apple Podcasts. You can also find more about Tom and Amodo's work at ai-verification.com/ 00:00:00 Using compute verification for an AI slowdown 00:06:08 Proving a datacentre is only running inference 00:13:22 Reading a GPU's workload from its power draw 00:23:35 What would it take to get verification deal-ready? 00:37:49 How realistic is verification? 00:46:07 Preparing for a rushed US-China deal 01:07:20 How good is AI at hardware design?

New on our Frontier Red Team blog: using Claude to formally verify software at large scale. Formal verification can prove that code is free of certain bugs, but historically it has only been feasible for small codebases (or required a lot of expert time). With AI models, however, this is beginning to change. In collaboration with @theoremlabs, we used Claude to prove security properties in over one hundred real-world open-source libraries over the span of a few weeks. Read our research in full here: anthropic.com/research/forma…

1/14) Virtual animals are a long-standing goal in neuro-AI, but they should actually behave like real animals. We present SCAMPER: a virtual rat for RL studies that moves naturally during both exploration, and goal-directed behavior. Project page: aidansirbu.com/SCAMPER-site/

Five major databases of human surface proteins agree on only 188 of them. So we had Claude read the literature instead. Today @deliverome is releasing an open atlas of the accessible human surfaceome, with claims traced directly to quotes from their source 🎯

Protein immunogenicity is a major driver of drug failure in the clinic, a problem especially important in the age of generative AI when we can design de novo medicines with desired function. Meet 🐐MESSI: a generative method for deimmunizing proteins with discrete diffusion. We reduce predicted MHC-II epitopes and experimentally validate retained or improved function in luciferases, PD-1 binders and uricases—an enzyme family used in FDA-approved therapies.💊 Watch the story below: 1/n 🧵

Functional recovery of vitrified and nanowarmed human-scale pig kidneys biorxiv.org/content/10.64898…

What would it take to fund ambitious science beyond the lab or startup? This autumn, we’re bringing people across the UK together to explore new models, including Focused Research Organizations (FROs). No prior knowledge needed. Join us, and bring a curious friend.

For a while now, @kirancodes has been excitedly telling everyone at @BasisOrg who would listen about a crazy idea she's been pursuing. Not everyone thought it was possible, but it's starting to work. See basis.ai/blog/ship-your-inte… TL;DR: The bottleneck in verifying the world's software is shifting from theorem proving to specifications and semantics. We now have a way to verify semantics that can scale across all programming languages. The basic idea: To verify that a program is correct, you need the program, a specification you want to prove, and a semantics that defines what programs and specs mean. Most programming languages (Python, Ruby, Lua…) have no formal semantics, because semantics are painful to write for complex languages. You can ask an LLM to write a semantics for you, but it'll probably have mistakes. But (i) programming languages do have interpreters/compilers, which ultimately are compiled down to machine code, and (ii) some instruction sets do have formal semantics. RISC-V has an official formal model. So, crazy idea: Step 1: Compile your interpreter (e.g. Python's) down to RISC-V. Now its behaviour is formally defined. It's exact, but unreadable: thousands of machine instructions. Step 2: Ask your LLM to write a semantics for your language. Step 3: Ask the LLM to prove, in @leanprover, that its semantics agrees with the compiled interpreter on every possible program. If the semantics is wrong, the proof gets stuck, and the LLM goes back to fix it. Step 4: If the proof goes through, you have a semantics you don't need to trust, and you can verify programs in a language that never had one, with guarantees that reach the binary that actually runs. Kiran has now done this end to end for a simple WHILE language, and we're scaling it up to Lua and OCaml, with Python as the north star.

I'm looking for technical scouts to find the world's best researchers whose dreams don't fit into academic lab- or venture capital-shaped boxes It is, first and foremost, an ideas job. Flexible on work arrangement. Technical vision must be first-rate. freaktakes.com/p/help-wanted…

“The @NSF is an engine of scientific discovery and innovation. It is an envy of the world.” @banksianr

Performance or longevity. For years, it seemed like the field of BCI would have to choose between the two. Our accumulating preclinical data suggests otherwise: four years of stable neural recording and decoding in large animal models, with steady spike SNR and no electrode array migration. Read the full announcement here: paradromics.com/news/paradro…