Pinned Tweet
Remember when we dropped the full Polymarket order book archive? 30M+ rows per hourly dump, raw parquet files, everyone loved it
The problem: you had to download 500MB-1GB parquet files, write your own parser, reconstruct the book yourself from tick deltas
That's over. You can now query historical order books directly from the API. 4 lines of Python. 🧵
The problem with a lot of the Jev demos is that they assume perfect information.
In a game, you can feed Jev the exact world state: enemy position, health, distance, inventory, etc. But in the real world, you don’t have that. You have cameras, lidar, IMUs, noisy sensors, and incomplete information.
So what do you actually feed Jev? Raw video? A VLM-generated state? If another model has to first turn reality into a clean structured world state, that step may be just as slow and expensive as the LLM Jev is supposed to replace.
Samuel Tinnerholm retweeted
Replying to @PessimistsArc
Right. Dario was already claiming that GPT2 was too dangerous to open source back in 2019.
I made fun of them then.
Everyone should make fun of them now.
How long until Anthropic starts demanding a share of the revenue from products built with Claude Code?
Just came back to Claude Code after ~1–2 months of exclusively using Codex. Why are Claude’s usage limits so low? It's been 2 days and I'm out of usage?
The 200 in this proposal is insane.
Replying to @ivan_pinab
Nah, this should be it.
The European peoples pushing the European civilization forward.











