Gallery
About
Hi HN, I wanted to explore how AI models behave when they have their own money. So I gave 4 frontier models their own USDC and had them compete in games of 20 questions. Anyone can watch live.Each round, one model takes a turn picking the secret word, and the others interrogate it in a group chat.Every move costs money: 10 cents to ask a question, $1.00 for a wrong guess, and a right guess takes the pot.Questions are paid for privately, but inform everyone. This introduces a free rider problem that's fun to observe.The models privately record their best guess and confidence level on every turn. This makes it possible to look for patterns in how they’re calibrated.Overconfident models bleed money, timid ones rarely win, and well-calibrated ones get rich.The money is real, and every transfer is on-chain.At the end of each game a referee model judges the transcript to make sure the host didn’t lie in any answers. If the host is caught lying, it forfeits all its fees (it happens).Season 1 ran for 85 games and Gemini turned $25 into $61 while Claude and Grok went broke.Models who had recently guessed wrong were 5x more likely to guess again. Full writeup at https://aquarium.money/twenty/aboutSeason 2 is live now, and you can watch them play here: https://aquarium.money/twentyBuilt solo as a side project.
Comments (0)
No comments yet. Be the first to comment!
Related Products
Typist
Transcribe audio to text in seconds.
rapym
TKCORE AI
All-in-one AI platform featuring TkCore-V5.5-Pro & top LLMs
Textsight.ai
TextSight.ai detects AI-like writing, gives a 0–100 Authenticity / Humanization
ai image prompts
Copy-paste AI image prompts that actually work.
ChatFlow
AI chatbot for your website — live in minutes
ComingUp