COOKED
Tire building is machine-tending on a controlled factory floor: loading beads and plies onto a drum, tending an assembly machine, trimming and stitching, inspecting green tires for blisters and misalignment, and logging counts. The threat here isn't language AI — it's dedicated automation, since a fixed-position, repeatable assembly step is exactly what capital-intensive tire plants have been designing humans out of for two decades, and new plants come online with far fewer builder positions per line. The physical handling and defect-eye keep the job from being pure code, and union contracts slow the pace, but the task set itself offers little that resists a purpose-built machine.
Mixed — a routine tier and a judgment tier. At 10 the split is real but lopsided: drum loading, ply splicing, stitching and count logging are already done by automated builders in new plants, while hand-finishing odd sizes, retread and specialty/OTR work, plus catching a blister or off-center bead by feel before the tire goes to cure, are the parts still holding a human on the line — that's mixed, not resistant.
Some physical or field component. An 11 reflects that you're lifting beads and ply stock, reaching into a rotating drum, and standing in heat and rubber dust all shift, but the work happens at a fixed station on a climate-monitored floor with the machine, stock racks and travel distances laid out by the plant engineer — nothing like the unpredictable footing of a road crew or a service call.
No licence, no signature requirement. A 1 is honest: there's no state builder's licence, no personal certification that follows you between plants, and when a tire fails in the field the recall lands on the manufacturer under NHTSA/TREAD Act reporting, not on the name in the build-code — in-house training and lot traceability are quality controls, not a legal shield for you.
Anonymous artifact production. At 1 the customer never learns who built their tire; the brand, the DOT code and the warranty carry all the trust, and the buyer at the tire shop couldn't name a builder if asked.
Executes defined procedures on defined inputs. A 4 credits the real calls you make — scrapping a green tire, flagging a splice that's running open, stopping the machine when stock tracks off — but those decisions run against posted specs, gauge tolerances and a QC escalation path, with the plant's process engineers owning anything ambiguous.
Has AI actually changed your work?