Over the past ten days, we completed assessments across six professions—financial planners, nurses, product managers, customer success leads, and others. The pattern that jumped out: not a single task description mentioned existing AI tool use, despite an average effective automation exposure score of 29%.
This doesn't mean AI hasn't reached these roles. It likely means something else: either these workers haven't yet integrated AI into their daily workflows, or they're not thinking of their tools as "AI" when they describe their work. (A financial planner using an AI-assisted research platform might just call it "research.")
The data is still thin—only ten new real assessments this period—so we're reading early signals, not conclusions. But the zero-adoption-claim against moderate automation exposure is worth tracking. If AI risk is real but adoption reporting stays near zero, it suggests a lag between technical capability and workplace integration. Or it reveals how we talk about tools: most people don't label their work by the technology behind it.
For readers in roles like product management or customer success—fields where language, judgment, and relationship work heavily—the 29% exposure score masks a bigger story. Those roles showed moderate to thick contextual knowledge requirements and social functions that create friction against simple automation. The real number isn't the headline; it's the gap between what *could* be automated and what *actually will be*, given how these jobs are really done.