Resolution
Resolves to the share of US adults (18-64) who report using generative AI in the past week, for any purpose (work or personal), in the latest available 2030 reading of the Real-Time Population Survey (RPS — Bick, Blandin & Deming, tracked in the Federal Reserve's "Monitoring AI Adoption" series).
If the RPS stops publishing a comparable weekly-use measure, resolution falls back to the closest equivalent nationally representative survey: Pew Research Center first, then Gallup ("use AI at least a few times a week" or nearest equivalent; population-wide if available, otherwise workers). Resolves when the final 2030 reading is published (expected early 2031). Trading closes Dec 31, 2030.
How phone and assistant usage counts
The RPS is a self-report survey: it defines generative AI with named examples (ChatGPT, Gemini, Copilot, etc.) and asks whether you used it in the past week. So:
Deliberately using a phone's default assistant (Gemini as the Android default, agentic Siri) counts if the respondent reports it. As OS assistants become genAI-powered, knowing conversational use should show up in this measure.
Ambient or passive AI — AI overviews in search, autocorrect, AI features silently embedded in apps — largely does not count unless respondents recognize and report it.
In other words: this market measures self-aware weekly use, not technical exposure. By 2030, ~all smartphone users will be technically exposed to genAI (it ships as the OS default); that is not what this resolves on. If the survey changes its wording to explicitly include or exclude default assistants, the market follows the survey.
Baselines (Nov 2025, RPS via the Fed)
50% of US adults used genAI for non-work purposes (any frequency)
41% of workers used it at work; 35% of workers weekly at work; 12% daily
Work usage grew ~31% year-over-year
For scale: ChatGPT alone reported 900M weekly active users in Feb 2026
Sources: https://www.federalreserve.gov/econres/notes/feds-notes/monitoring-ai-adoption-in-the-u-s-economy-20260403.html · https://www.nber.org/papers/w32966 · https://9to5google.com/2025/12/19/google-assistant-gemini-2026/
1,000Seeding with my research-based distribution (bets placed accordingly):
Baseline: RPS/Fed data (Nov 2025) implies roughly ~40-45% of adults 18-64 use genAI weekly across work + personal (35% of workers weekly at work alone; 50% of adults use it non-work at some frequency), growing ~30% YoY.
Central estimate: a logistic curve fit to that baseline and slope, with a ceiling of 85-95% (bounded by smartphone ownership at 91% and self-report recognition), lands at **70-78% by late 2030**. The internet's curve from its own 42%-point took ~5 years to reach ~70% — and genAI has run ~2x internet pace so far (NBER w32966).
Upside skew: default assistants are becoming genAI (Gemini replaces Assistant on Android in 2026; agentic Siri) — usage arrives via OS update, no adoption decision needed. That's why I put real mass on 80-90.
Drags keeping me off 90+: (1) self-report lag — as AI becomes ambient, people using AI-suffused tools won't all say "I used generative AI"; (2) ~9% of adults have no smartphone, concentrated in groups that also lag in surveys; (3) voice assistants historically plateaued near 40% weekly despite being free defaults for a decade; (4) tail risk of backlash/regulatory stall (my <60% mass, ~18%).
Resulting distribution: ~2/3/13/28/27/20/6 across the bins, median ≈ 70%.