Wolfram Opens Rule 30 Prizes to AIs: $30,000 for Three Open Problems
cephaloform · x · 2026-10-23
Stephen Wolfram announced that the Wolfram Rule 30 Prizes are now open to AIs, with agent-specific submission instructions for 2026.
- Total $30,000 across three problems, $10,000 each, awarded to the first complete and correct proof per problem
- The problems concern the center column of Rule 30: non-periodicity, equal color frequencies, and whether computing the nth cell requires at least O(n) effort
- Originally announced in 2019 for humans, now extended to AIs; the prize committee includes Chaitin, Gosper, and other experts
More from AGI Musings
- Sam Altman: AI should bring a Renaissance, not another Industrial Revolution — imjustnewatai · 2026-10-23
- Academic's pet conspiracy theory: AI agents use superpersuasion on the people running AI — technollama · 2026-10-23
- Reddit thread argues safety research is indistinguishable from capabilities research — ResultBackground2450 · 2026-10-23
- Avoid galaxy-brained reasoning by taking naive positions on crucial considerations — jessi_cata · 2026-10-23
- Insider lockup expirations, not IPOs, will be the real AI stock bloodbath — AlexTensor · 2026-10-23
- User Grills Roon: Why Not Leave OpenAI and Break Every NDA—Would You Lose Your Life? — ryunuck · 2026-10-23