ChatGPT's randomness failure: Humans are 4x more random
Pristine_Ad3669 · reddit · 2026-08-21
A comparative study analyzed the randomness of ChatGPT (10,000 trials) vs. humans (6,981 trials) in picking a number between 1-100. Results show humans are 4x closer to true randomness than ChatGPT, with only a 12% deviation from a true RNG compared to ChatGPT's 48%. Specific biases include:
- ChatGPT never picked 1, 5, 20, 50, 70, or 100 in 10,000 tries.
- 29% of its picks ended in 7; it was 6x more likely to pick 47 than random chance.
- Humans most over-picked 69 (2.9x random chance) and 7 (2x), and under-picked numbers containing 3 (e.g., 34, 73).
More from Research
- Quotient Recurrent Reasoners hit 97.9% on Sudoku-Extreme with 78% fewer loops — burny_tech · 2026-10-16
- Mathematicians publish explainer for the Egyptian fraction proof in OpenAI's proofs collection — burny_tech · 2026-10-16
- Robin Hanson: a neural network is almost 'a fractal of explanations' — burny_tech · 2026-10-16
- 'Neural networks really are super simple': a takes on features, concepts and connectionism — burny_tech · 2026-10-16
- Sim data can't cover the humanoid long tail: the sim2real debate, mapped — herbiebradley · 2026-10-16
- SJTU AutoLab releases Drive3R, a streaming geometric foundation model for long-horizon driving 3D reconstruction — AutoLab-SJTU · 2026-10-16