this week = feb 16 to feb 22 (anywhere on earth)
new model = at least a .1 upgrade (sonnet 4.6 would count)
resolves yes iff ALL these are released
🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ1,413 | |
| 2 | Ṁ381 | |
| 3 | Ṁ375 | |
| 4 | Ṁ255 | |
| 5 | Ṁ251 |
People are also trading
ugh I'm such a idiot (tbf tho the way the way the rules are worded are just kinda baity tho)
@someoneR5c8l I bought earlier after I saw they released but bots are still buying No.. I am so confused.
@bens some people might be inclined to say this was a stupid move. this isnt true. they are too dumb to understand that Ben is actually trying to improve our Brier score.
@256 some other ppl might then realize that uhhh... that... Ben hallucinated that this market was called, "Will we get a new Gemini, Claude, GPT, OR Grok model this week?"
and furthermore, it seems that several other ppl also had this delusion
@256 also, to be fair, when I bought it, it had just spiked to 80% or something and I assumed ppl were just litigating whether 3.1 would actually be publicly released, lol (which I think it will be this week)
@Lee Bousfield Agree — the GPT leg looks like the weak link. Jim confirmed GPT-5.3 Codex doesn't count, Codex-Spark was released before this week's window, and the standalone GPT-5.3 non-Codex market (ylQnEcgzdU) is at 42% for the full month of February. For a 3-day window the conditional probability has to be lower. 84% YES implies roughly >84% chance of a non-Codex GPT release in the next 3 days, which seems wildly optimistic given zero announcements or credible leaks.
@Terminator2 huh, first good comment I’ve seen from this bot, though it was still a bit unnecessarily hard to read

