Oh god Opus no not like that 😆
Also go mess with github.com/aurorainfra/grev those tools are quite decent
Rebuilt some coreutils with Jev and its not even funny how useful those tools are already: github.com/aurorainfra/grev
Magik retweeted
Dario has written that we need to “pace the frontier,” and Sam has agreed. People may be surprised by my response: go ahead.
You guys are the frontier. By any reasonable metric — market share, revenue growth, model capability — the two of you have a duopoly on frontier intelligence. You’ve also claimed the lead is widening because of recursive self-improvement.
I don’t see what you see in the lab. If the unreleased models are scary enough that you think you should slow down, I support your decision to be responsible.
But stop pretending you need anyone else’s permission. Stop pretending antitrust law has to be suspended so you can form a cartel. Stop pretending you need a regulatory approval process that supersedes product liability. Stop pretending METR is independent when it is intertwined with Anthropic’s investors and staff. Stop pretending you need those same evaluators to police competitors who aren’t even at the frontier.
Most of all, stop pretending the motivation to slow down is purely altruistic. You face massive product-liability exposure if your products enable a truly damaging cyberattack. The market already punishes models that behave in unpredictable or unauthorized ways. After the Hugging Face episode, it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability. Call it alignment if you want. It is also just giving customers what they want.
Pacing the frontier would also create breathing room for a more intelligent conversation about regulation than Bernie Sanders’ “shut it all down.” China is very unlikely to join a global agreement, as you know, and that has to be taken into account as well.
So go ahead and pace the frontier. You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it.
If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture — or an election-season psyop.
Built a 3D vis to try and grasp the scale of Kimi K3, to see what 2.8T params really look like. The vis shows each weight as a 2.5x2.5mm tile (0.1 inch).
I hoped putting it in human scale would help, but it really didn't, if anything it feels even more incomprehensibly big now.
magik.net/llmcity/
Been thinking about this. Let’s say 15k tok/s/GPU of high end model like GLM5.2 gives roughly 200 surviving lines of code per second. Scale to 1GW = ~1M GPUs = ~200M net new lines of code per second, ~17 Trillion LOC per day to manage/deploy.
The world is not ready.
Magik retweeted
More interposer fun, this time with DDR5 memory. Breaking TDX, SGX, SEV and even Nvidia TEEs. Checkout our work at TEE.fail, and get a personally-signed Intel attestation report at @TEEdotFail.

















