@jim Is "open weights" is sufficient to count for this?

Accord to NYT? Never. Highest Eli score would b due to “copying”, “trickery”, “undisclosed info in the 1st page of disclosure doc”, “conspiracy”…besides “we don’t use Elo anymore”.

Next article “China threatens the world with leading LLM model”

This market splits the manifolders in half

hey @Bayesian don't be pusillanimous, place your bets!

@jim sorry what, the probability of this is like 10%

oh, open model. nvm

hmmm yeah i'll buy this up

@Bayesian ya I had the same reaction... I still think there's less than 40% chance of this by EOY tho, tbh

ig google might cook up some decent OS stuff

as might meta

arbitrage with https://manifold.markets/Sss19971997/best-open-weight-llm-by-eoy-2025-by