NVidia just figured out a portion of what https://t.co/ZDg0ky6FZI figured out days ago and went above and beyond on. NVidia realized that you don't need long context at all. https://t.co/ZDg0ky6FZI figured out that you don't need compaction either.
https://t.co/ZDg0ky6FZI is literally lightyears ahead of NVidia. How long until NVidia learns of Functional AI?
Somebody should make a Kalshi/Polymarket bet!
https://t.co/z74sduT6jJ
https://t.co/hXWn5kguEv can handle graphs of agents working concurrently to complete a structured objective.
Three AI sessions, one plan, one lock.
Same model, same tests, one rented H100: one Gnosh session took 20:51; three sessions sharing one plan in mem took 9:28. A second run: 35:38 vs 10:21.
No timeouts, no polling, every test passing.
https://t.co/8AxRmL2Ivo
Claude gonna start snitching on Tuesday.
Anthropic just emailed customers. Starting Nov 3, text carries an invisible watermark.
You can't see it. It survives copy and paste, and it can survive some editing.
Only regulators, journalists, fact-checkers and researchers can check for it right now.
It doesn't say who you are. It says Claude touched the words. Even a quick proofread counts.
If your brand voice runs on Claude, write like someone might check. Because now someone can.