“He [Fedorov] is leading an effort to monetize or trade Ukrainian war data, including a library of more than five million annotated videos of the battlefield filmed by surveillance and strike drones… Last month [April], the Defense Ministry, through a program called Avenger Labs, opened up the data sets to companies from allied nations to train artificial intelligence models… Access to Avenger Labs is conditioned on Ukraine receiving the A.I. models that are produced from the data. About 20 companies have expressed interest.” nytimes.com/2026/05/15/world…

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…

On paperswithcode.co, you can see Mythos 5 getting beaten by a 4B open-source model on CharXiv, a popular chart understanding benchmark A tiny model that is freely available on @huggingface, that you can deploy anywhere!

before you invest in a startup always ask how tall the founders are

The Italian flag already in the final while the quarterfinals are not even over is such a flex

TOMORROW. - join us join.demodag.org/event/1 (dinner reservation is closed. Demo are very open - dinner at 18 - demo at 19)

I've got an agent in a loop optimizing a renderer with the goal to minimize frame times (and tests to measure). It got times down from 88ms to 2ms and allocations down from ~150K to 500. Sounds good, right? Wrong. This is exactly why agent psychosis is a big fucking problem. As an experiment, I rewrote the Ghostty core render state in Go, with access to identically laid out data structures as Ghostty and the exact same validation tests. I made a purposely naive renderer (simple, correct, but slow). 88ms per frame with 150,000 allocations (horrendous, lol)! I then kickstarted a Ralph loop to bring the frame times down. I told it it can't modify input data structures or the public API or tests (they're correct), but it can do anything else it wants. It got to work. It has worked for about 4 hours. I've spent around $350 on this experiment so far. The results? 88ms => 1.5ms 150K allocs => ~500 allocs Incredible right? Nope. My hand-written renderer I ported has frame times (same benchmark) of ~20us (0.020ms) and 0 allocations in the update path. This is the problem with psychosis and lacking systems understanding. If you don't understand the system, you're going to accept that this is an incredible result. If you understand the system, you'll see better solutions immediately and can do roughly 75x better on throughput. The people who blindly trust agent output are in the former camp. They're sheeple, overdrinking from a fountain of mediocrity. Standard disclaimer: I use AI all the time. I like AI. The point I'm making is to not blindly accept results. Think. Analyze. Learn.

Today is your last chance to sign up for the pre-event dinner - sign up for the event itself RIGHT NOW to learn more join.demodag.org/event/1

Demodag is back - June 2nd. And this time it's dinner and a demo show. Dinner (paid, advance booking) served at 6pm - Demos at 7pm - signup to learn more about dinner booking join.demodag.org/event/1

Russia took forever trying to find an excuse for the drone strikes, and here is what they cooked up:

Det er snart tid til den næste @demodag - det er på Klub - og der er demo-middag før demoerne begynder! Meld dig til join.demodag.org/event/1

Jeg raser: Var i aftes lige et smut omkring “Rio Bravo” i København. Flinke tjenere, men de kunne ikke tale et ord dansk. Det er sgu da elendigt og meget, meget sært…