Will any LLM produce a reasonable poker simulation, as judged by Nate Silver, by the end of 2028?

1kṀ1585

2028

54%

chance

ALL

https://open.substack.com/pub/natesilver/p/chatgpt-is-shockingly-bad-at-poker

I'm expecting Nate to repeat this test. If he doesn't even try, this resolves NO.

Update 2025-05-22 (PST) (AI summary of creator comment): The creator has specified conditions regarding tool usage by the LLM, in line with Nate Silver's post about testing for general intelligence:
- The LLM must not use tools specifically designed for poker (e.g., dedicated poker solvers).
- The LLM may use generic tools, such as the ability to write and execute Python scripts.

Update 2025-05-23 (PST) (AI summary of creator comment): The market may resolve YES if the AI performing the task is considered the 'main frontier approach to general purpose AI', even if it is not strictly classified as an LLM. This is contingent on the AI not using techniques similar to a traditional poker solver.

Technology

Politics

OpenAI

Technical AI Timelines

Get

1,000

to start trading!

People are also trading

In 2025, will I be able to play Civ against an LLM?

15% chance

Will RL work for LLMs "spill over" to the rest of RL by 2026?

35% chance

Will an LLM consistently create 5x5 word squares by 2026?

84% chance

Will the best public LLM at the end of 2025 solve more than 5 of the first 10 Project Euler problems published in 2026?

75% chance

Who will have the best LLM at the end of 2025 (as decided by ChatBot Arena)?

Will any LLM be able to consistently play Akinator correctly as the user by 2028?

79% chance

Will there be a state-of-the-art LLM that is NOT based on next raw token prediction before 2029?

50% chance

What organization will top the LLM leaderboards on LMArena at end of 2025? 🤖📊

Will LLMs be able to formally verify non-trivial programs by the end of 2025?

27% chance

Will any LLM be able to multiply together arbitrary decimal numbers by the end of 2027?

Sort by:

How does this resolve if the "main frontier approach to general purpose AI" is no longer considered to be an LLM but is capable of doing this?

@rayman2000 I would probably count that as long as it doesn't use the same techniques as a traditional poker solver.

bought Ṁ500 YES

is the ai allowed to use poker solvers as tools to make the decisionmaking more realistic?

@Bayesian No, based on Nate's post, the idea is to use poker as a test for whether an AI has general intelligence:

"What would be more impressive — more of an indication of general intelligence — is if large language models like ChatGPT that weren’t designed for poker could converge on a good strategy just by crunching textual data and learning to reason from it. But for now, they aren’t close to doing that well."

I'll allow using generic tools like writing Python scripts, but not anything specifically related to poker.