EXPOSURE TO AI
18%
RESILIENT
OBSERVED IN REAL USE · Anthropic 2026
0%
of this role’s work is already showing up in real Claude usage (Anthropic Economic Index).
The last desk the machine reaches. Breathe.
18% 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. Certified hard to automate. Today’s AI barely touches the core of this one.
WHAT AI CAN ALREADY DO
- Explain how various menu items are prepared, describing ingredients and cooking methods
- Write patrons' food orders on order slips, memorize orders, or enter orders into computers for transmittal to kitchen staff
- Present menus to patrons and answer questions about menu items, making recommendations upon request
- Inform customers of daily specials
- Describe and recommend wines to customers
WHAT IT STILL CAN’T
- Take orders from patrons for food or beverages
- Check patrons' identification to ensure that they meet minimum age requirements for consumption of alcoholic beverages
- Collect payments from customers
- Remove dishes and glasses from tables or counters, and take them to kitchen for cleaning
- Serve food or beverages to patrons, and prepare or serve specialty dishes at tables as required
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 18% 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