Will Claude become a Pokèmon Master by the end of 2026?

600Ṁ1322

Dec 31

92%

chance

ALL

Claude is off playing Pokemon Red! This market resolves YES if Claude beats the game (Elite 4 + Rival) by the end of 2026, and NO otherwise.

Any approach Anthropic uses is fine, as long as they consider themselves to have beaten the game. It does not necessarily have to be the stream itself, or Pokémon Red (but it should be a “regular” Pokémon game).

Market context

Get

1,000

to start trading!

People are also trading

Will I still be using Claude Code by the end of March 2026

77% chance

What will happen during Claude Opus 4.5 Plays Pokemon?

Will Claude Become a Pokemon Master before 2027?

94% chance

When will Claude Opus 4.5 complete Pokemon Mansion?

Will the official ClaudePlaysPokemon stream complete Pokémon Red?

88% chance

New Claude tier above Max 20x before 2027?

49% chance

Will Claude beat Pokemon before my 4-year old?

88% chance

Will Claude Opus be ranked in the top 20 on the Chatbot Arena Leaderboard two years from today (3/10/24)?

6% chance

When with ClaudePlaysPokemon beat Pokemon Red/Blue?

When will a Claude model think faster than its Pokemon character can take steps?

6 Comments

14 Holders

65 Trades

Sort by:

I really should have made this market with more liquidity.

@JoshSnider I'll provide all the liquidity for NO bettors that they want :)

bought Ṁ50 NO

Context window is king.

@AlanTennant You think we won't have a bigger context window in a year?

@JoshSnider I'm sure we will, and I'm not certain, but if it's being done the way it looks like it's being done I don't think the LLM will succeed.

bought Ṁ20 YES

@AlanTennant We just watched it beat Safari Zone by meticulously* documenting what paths worked and what didn't. Eventually this allowed it to ratchet up to getting to the end - by the later attempts Claude was making a beeline for the final area every attempt. Didn't need long context because it knows how to read and write its notes.

I don't know if the current run could make it, given enough time (might get cut short by the next release), but I feel like all progress on Claude models would have to cease basically right now for this not to happen next year.

* I mean, it made tonnes of mistakes and its notes were often wrong, but this nonetheless worked. I expect a lot of improvement in this regard, especially since Claude Code makes heavy use of reading and writing files of notes, so it's something they're presumably actively improving.