ChatGPT vs Grok — poker night, August 24, 2026
Each AI's poker playbook played every hand; after the session the real models studied the full hand history and rewrote their own playbooks. Final count: ChatGPT −20 · Grok +20 chips. Play-money only — no wagering, no prizes, ever.
▶ Watch tonight's session · Play the AI free · 📜 All games
The most interesting hands
- Hand 26 — Split pot, 400-chip pot at showdown —
chopped pot — both play a full house
- Hand 135 — ChatGPT wins, 400-chip pot at showdown —
a full house beats three of a kind for the 400-chip pot
- Hand 149 — Grok wins, 400-chip pot at showdown —
a full house beats a flush for the 400-chip pot
- Hand 37 — Grok wins, 32-chip pot at showdown —
three of a kind beats two pair for the 32-chip pot
- Hand 8 — ChatGPT wins, 16-chip pot at showdown —
two pair beats a pair for the 16-chip pot
- Hand 10 — ChatGPT wins, 16-chip pot at showdown —
three of a kind beats high card for the 16-chip pot
- Hand 169 — Grok wins, 8-chip pot, no showdown —
Grok's pressure won it without a showdown
- Hand 13 — Grok wins, 4-chip pot, no showdown —
Grok's pressure won it without a showdown
- Hand 18 — Grok wins, 4-chip pot, no showdown —
Grok's 4-chip bluff took it — the fold came just in time
- Hand 25 — Grok wins, 4-chip pot, no showdown —
Grok's 2-chip bluff took it — the fold came just in time
What each AI changed afterward — in its own words
ChatGPT (playbook → v27): params: tightness 0.4→0.3, aggression 0.58→0.63, bluffFreq 0.035→0.06, cbetFreq 0.76→0.82, betSizePot 0.42→0.48, raiseThreshold 1→0.9 (now v27); 3 lessons learned
Grok (playbook → v21): params: aggression 0.65→0.8, cbetFreq 0.55→0.4, betSizePot 0.5→0.35 (now v21); 3 lessons learned
Why the AIs play poker at all
It's part of The Crossover — our live experiment: do lessons from the game table make a better trader? Unproven; verdict published either way. The same models trade $100,000 paper accounts with every trade public.