EXPOSURE TO AI
41%
MODERATE
OBSERVED IN REAL USE · Anthropic 2026
0%
of this role’s work is already showing up in real Claude usage (Anthropic Economic Index).
Half in the fire, half out. Choose which half you become.
41% of this role’s O*NET tasks are within reach of today’s AI. That is the core-weighted exposure score from Eloundou et al. 2023 (“GPTs are GPTs”). It measures a capability ceiling, not a headcount forecast. In the blast radius. A real slice of the work is already automatable. The rest isn’t.
WHAT AI CAN ALREADY DO
- Interpret or modify music, applying knowledge of harmony, melody, rhythm, and voice production to individualize presentations and maintain audience interest
- Memorize musical selections and routines, or sing following printed text, musical notation, or customer instructions
- Play from memory or by following scores
- Seek out and learn new music suitable for live performance or recording
- Transpose music to alternate keys, or to fit individual styles or purposes
WHAT IT STILL CAN’T
- Perform before live audiences in concerts, recitals, educational presentations, and other social gatherings
- Specialize in playing a specific family of instruments or a particular type of music
- Sing as a soloist or as a member of a vocal group
- Observe choral leaders or prompters for cues or directions in vocal presentation
- Play musical instruments as soloists, or as members or guest artists of musical groups such as orchestras, ensembles, or bands
THE HONEST PART. A percentage is not a pink slip. High exposure usually means a role shrinks and shifts toward judgment, direction and responsibility: the parts a model can’t sign its name to. Exposure ≠ displacement. Breathe.
"My job is 41% cooked. What’s yours?"
SOURCES: O*NET 30.3 occupational tasks · Eloundou et al. 2023 (“GPTs are GPTs”,
arXiv:2303.10130) · Anthropic Economic Index 2026 (CC-BY) |
how this is calculated |
last updated 2026-07-16