The model quality just crossed the threshold. Claude Sonnet can now produce genuinely useful, voice-matched outputs on real professional tasks — and web search plus computer use means it can retrieve live information and act in the world, not just generate text.
The people being displaced are the highest-value underserved segment. Experienced knowledge workers leaving long-held roles have deep careers, urgent need, and no structure to run a transition with. They don't want another job board or a coaching portal. They want something that already knows their history and does the work. Nobody has built that.
Employers already fund this moment — and get almost nothing for it. 87% of HR leaders ran or plan layoffs in 2026, driven by skills displacement, AI transformation, and shifting market demands. 77% say they offer targeted redeployment and mobility programs; only 19% of employees recognize them. The budget exists; the outcome does not. And nobody in the category publishes an engagement figure for outplacement at all — which tells you what is not being measured.
The platform giants are building engines, not relationships. OpenAI, Anthropic, and Google compete on model capability. None of them arrive the week someone loses a job, learn that person's career, and stay for the decade after. Distribution at the moment of maximum motivation is not a model problem.