A decade after the ‘Godfather of AI’ said radiologists were obsolete, their salaries are up to $571K and demand is growing fast

The market takeaway is not that AI eliminates a labor category; it is that in regulated, liability-heavy workflows, AI mostly reallocates human time to the bottleneck tasks that are hardest to automate. That preserves headcount demand while improving throughput, which is constructive for infrastructure spend and less so for any investor thesis built on rapid labor destruction. In public equities, that argues for continued capex into compute and workflow software rather than an abrupt collapse in service demand.

Second-order, the bigger economic effect may be volume expansion: when automation lowers friction, utilization rises, and the scarce resource becomes the human reviewer or procedural specialist. That is structurally positive for imaging-related equipment, storage, and inference layers, but it can also keep wage inflation sticky for labor-intensive providers. The losers are companies that need an immediate, visible productivity step-up to justify margin expansion; the article weakens the case that AI alone can deliver that inside one or two earnings cycles.

The contrarian miss is timing: consensus often extrapolates task automation into company-level headcount cuts, but reimbursement, malpractice, and workflow integration slow the translation. Unless regulators change the final-read requirement or liability standards, full substitution stays a years-long outcome, not a quarter-to-quarter catalyst. The falsifier is simple: if CMS/FDA rules loosen or management teams begin showing sustained, measurable opex compression from AI in the next 2-4 quarters, the current 'augmentation not replacement' thesis needs to be revised.