DIY UBI math with real data: about $2,000 per person after 30 years, not a living wage
edmrunmachine · reddit · 2026-11-01
The author ran a serious back-of-envelope analysis of UBI in America using AI assistance and real company data:
- The plan: "everybody owns a piece of everybody's robot" — the public gradually acquires 25% of the 500 biggest companies, dividends are paid out equally and locked in.
- Four rounds of cross-checking across Claude Opus 5.5, GPT Astra 6.0, Gemini Argon 4, and Muse Wonder, each catching mistakes others missed.
- The result: about $2,000 per person per year in 30 years, maybe $6,000 in an AI boom — a stimulus check, not a living wage.
- The hard constraint: even 100% public ownership of the top 500 equals only $4,600/person today; all US corporate profits are $3.7T/year pre-tax, while a $20,000/person UBI would cost $6.8T — nearly double all corporate profit.
Code and full write-up are public (GitHub + Substack); the author invites people to tear it apart.
More from AGI Musings
- Burkov: humans validating machine work is no different from using a tractor — burkov · 2026-11-01
- Octant and Schrödinger profile 700+ molecules to explain why pKa and logP predictions clash with experiments — _onionesque · 2026-11-01
- OpenAI staffer predicts full recursive self-improvement within 18 months — willdepue · 2026-11-01
- Gary Marcus meets Pope Leo XIV and agrees to serve as his AI ambassador — we_are_mammals · 2026-11-01
- Codie Sanchez: AI makes mid performers unnecessary — raise owners, not employees — irinarish · 2026-11-01
- Prediction: Next year is the year of AI-run autonomous online businesses — curious_vii · 2026-11-01