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

Augmentation vs automation share (Anthropic Economic Index)

consistent with normal60/95 confidence, mixed or hard to operationalisegrade Bcoincident

1What this measures

Share of classified Claude.ai conversations in the augmentation bucket (learning, task iteration, validation) versus automation (directive, feedback loop), computed from Anthropic's Economic Index open dataset for each 2026 release; the blog-reported shares and the share of occupations with at least a quarter of their tasks used are shown alongside.

Why it matters. The paper's bet is that AI diffuses as a tool people use, not a replacement; this is the only large-scale usage classification that speaks to it directly.

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

2How we track this

  • series anthropic_ei.global.augmentation_share.pt
  • series anthropic_ei.global.automation_share.pt
  • series anthropic_ei.global.augmentation_share_reported.pt
  • series anthropic_ei.global.automation_share_reported.pt
  • series anthropic_ei.global.occupations_25pct_tasks_share.pt
  • source Anthropic Economic Index (open dataset and reports) · default tier 2 · CC BY 4.0 (dataset)
Normal band
≥ 50.0%
Fast band
≤ 40.0%
Falsifying
—

Normal = augmentation at least half of classified use; fast = automation clearly dominant (augmentation under 40%). Between is `emerging`.

Applied to anthropic_ei.global.augmentation_share.pt.

3Tracker interpretation

Augmentation has held just above half of classified use through the 2026 releases (53% in Nov 2025, 54% in Feb 2026, 51% in Apr-May 2026); automation is gaining slowly on the monthly data.

4Evidence

Latest point
51.4%as of 2026-05-31
global
Value the bands apply to
51.4%as of 2026-05-31

11 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

Anthropic Economic Index, January 2026 report (November 2025 data): 52% of Claude.ai conversations classified as augmentation, 45% automation; the March 2026 report says augmentation rose again but only in figures. Inside the normal band (augmentation >= 50%). Tier 2 provider disclosure. Initial seed.

6Timeline notes

  • 2026-05-31 global · 51.4%as of 2026-05-31
  • 2026-04-30 global · 51.0%as of 2026-04-30
  • 2026-02-11 global · 54.4%as of 2026-02-11
  • 2025-11-19 global · 53.3%as of 2025-11-19

7Counterevidence

What cuts against this reading

One provider's consumer product; classification is model-based; API usage (where automation concentrates) is reported separately and skews the other way.

8Update history

  1. 2026-09-10unmeasured to consistent with normalconf — → 60 · claude-initial-seed

    Anthropic Economic Index, January 2026 report (November 2025 data): 52% of Claude.ai conversations classified as augmentation, 45% automation; the March 2026 report says augmentation rose again but only in figures. Inside the normal band (augmentation >= 50%). Tier 2 provider disclosure. Initial seed.

    obs:713109bb

9Confidence

60 / 95 — mixed or hard to operationalise

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