Skip to content
Slow Variables

Ask the data

Answers come from the same store as the site; every number is checked against the record it cites.

Early adoption · Adopters

Share of work hours assisted by generative AI

consistent with normal80/95 confidence, good evidence, some ambiguitygrade Bcoincident

1What this measures

Share of all US work hours in which employed adults used generative AI, from the Bick–Blandin–Deming Real-Time Population Survey (quarterly series via FRED), with weekly work use and any-use shares shown alongside.

Why it matters. The intensive margin. Headline adoption is broad, but hours actually assisted is what productivity arithmetic runs on.

Proxy types
behaviour, deployment
Unit
share
Cadence
quarterly
Valve
product to adoption

2How we track this

  • series fred.us_workers.hours_assisted_share.q
  • series fred.us_workers.work_use_weekly_share.q
  • series fred.us_adults.any_use_share.q
  • series bbd.us_workers.hours_assisted_share.pt
  • series bbd.us_workers.work_use_weekly_share.pt
  • series bbd.us_adults.any_use_share.pt
  • series stlouisfed_blog.us_workers.hours_assisted_share.pt
  • source Bick–Blandin–Deming Real-Time Population Survey (via FRED) · default tier 6 · FRED terms of use; RPS by Bick, Blandin & Deming
  • source FRED (Federal Reserve Economic Data) · default tier 6 · FRED terms of use; RPS series by Bick, Blandin & Deming
  • source FRED Blog · default tier 6 · FRED terms of use
Normal band
≤ 9.0%
Fast band
≥ 20.0%
Falsifying
—

Normal = single-digit share of hours in year three (PCs and the internet were here at the same age); fast = more than 20% of hours assisted.

Applied to fred.us_workers.hours_assisted_share.q.

3Tracker interpretation

Broad but shallow, exactly as the paper predicts - most workers touch it, few hours run on it. The weekly-use share moved far faster than the brief expected.

4Evidence

Latest point
6.3%as of 2026-06-30
us_workers
Value the bands apply to
6.3%as of 2026-06-30

27 observations. Hollow points are disputed (see counterevidence). Every point links to its observation.

5Status and reasoning

consistent with normalsince 2026-09-10 · claude-initial-seed

Bick-Blandin-Deming RPS, May 2026 wave via FRED: 6.3% of work hours assisted by generative AI, inside the single-digit normal band. Weekly work use is 39.2% and any-use 61.8%, both well above the brief's expectations - broad adoption, shallow intensity. Initial seed.

6Timeline notes

  • 2026-06-30 us_workers · 6.3%as of 2026-06-30
  • 2026-03-31 us_workers · 6.0%as of 2026-03-31
  • 2025-12-31 us_workers · 5.6%as of 2025-12-31
  • 2025-09-30 us_workers · 5.7%as of 2025-09-30
  • 2025-06-30 us_workers · 4.9%as of 2025-06-30
  • 2025-03-31 us_workers · 4.7%as of 2025-03-31

7Counterevidence

What cuts against this reading

Self-reported survey; the instrument changed between waves; hours assisted is not hours saved.

8Update history

  1. 2026-09-10unmeasured to consistent with normalconf — → 80 · claude-initial-seed

    Bick-Blandin-Deming RPS, May 2026 wave via FRED: 6.3% of work hours assisted by generative AI, inside the single-digit normal band. Weekly work use is 39.2% and any-use 61.8%, both well above the brief's expectations - broad adoption, shallow intensity. Initial seed.

    obs:5dcc48a7

9Confidence

80 / 95 — good evidence, some ambiguity

Confidence is independent of status: 90–95 multiple strong independent sources; 70–89 good evidence, some ambiguity; 50–69 mixed or hard to operationalise; below 50 limited or vague.

10Related