AI exposure: Bicycle Repairers

At 14.6% of weighted task load, Bicycle Repairers sits at the 23th percentile, below the point where a job's centre of gravity has moved. 78.8% of what this role does is untouched, meaning current systems cannot produce that work at all, whatever the commercial incentive.

What holds the line here is embodiment. Across this occupation's 18 tasks it averages 2.41 out of 3, the highest of the five friction dimensions. In plain terms, the work has to happen in physical space. A language model cannot move matter. Until the robotics to do this work is both good enough and cheap enough to deploy widely, capability in software does not reach it.

The most exposed thing this job does is Order bicycle parts, at 80.0%. The least is Weld broken or cracked frames together, at 0.0%. A gap of 80.0% between two parts of the same job is the reason this index publishes at task level. An occupation-wide number would have hidden both.

Within installation, maintenance and repair occupations, this one is more exposed than most. The median across the 50 roles in the group is 12.8%, and only 21 of them score higher than this. Occupational families are not uniform, and the spread inside them is often wider than the gap between them.

What would move this score. Of 18 tasks, 4 are currently banded exposed, 0 assisted and 14 untouched. For that distribution to shift materially would take robotics cheap and reliable enough to deploy at scale, not a better language model. The score is re-computed every quarter against a fresh capability reference, and the change is published rather than quietly applied.