Benjamin Todd: dumb trend extrapolation gets to AGI by 2030, no RSI needed
ben_j_todd · x · 2026-11-15
80,000 Hours founder Benjamin Todd argues that simply extrapolating the last four years' trends reaches broadly human-level 'digital worker' AGI in 2-4 years.
Two key trends:
- Revenue: OpenAI + Anthropic exceeded $115B annualized as of Sept 2026, tripling yearly for 4+ years (2024: 3x, 2025: 4.7x, 2026: 6x and still accelerating). Four more years of that trend implies $10 trillion — roughly 10% of world GDP, matching an estimate of AI absorbing remote/white-collar work (30% of GDP in wages) at lower cost. Revenue is his favorite benchmark because it's hardest to fake.
- Task length: AI task horizons have grown 8x/year since 2024 (doubling time shortened from 260 to 92 to 56 days), reaching multiyear tasks in early 2030.
He concedes 'every exponential becomes an s-curve' but notes revenue is still accelerating, so the burden of proof is on showing progress stops. If RSI has begun, superhuman AI could arrive in 12-18 months.
More from AGI Musings
- PhD research proposes 'intra-turn interaction': AI UIs shouldn't be turn-based — mattbeane · 2026-11-15
- AI task-horizon doubling now 56 days: trend extrapolation yields AGI by 2030 — ben_j_todd · 2026-11-15
- US voters rejected data centers 44 of 45 times; China builds a million-chip AI campus — ccerrato147 · 2026-11-15
- Gary Marcus: Doomers who can't tell an LLM from Claude Code are arguing in a muddle — GaryMarcus · 2026-11-15
- Ben Todd: 3 factors explain AI hitting 10,000-hour math tasks without superexponential growth — ben_j_todd · 2026-11-15
- MotherDuck CEO: AI agents will collapse data teams into one Data Intelligence Engineer role — teodorio · 2026-11-15