Devil's advocate: are we just building a more sophisticated form of credentialism?

Honest answer: nothing structural is different yet. The framework's better. The deployment incentives aren't. If we don't deliberately design against vendor capture and selection-effect drift, you're right, this ends in the same place.

From the buyer side: the thing that would actually be different is open-book scoring (anyone can audit how a score was computed) and per-applicant transparency (the test-taker sees what the score represents). Both are technically possible and politically hard.

Carmen's right that the historical track record is bleak. The cynical-but-honest framing: the framework gives us better tools to argue against the same outcome, not a guarantee of a different outcome. That's still a step forward.

30 years in federal hiring — Carmen, your skepticism is earned. The piece that's different now is that the cost of building competing assessments has collapsed. The capture story requires high-cost incumbents. We don't have those any more.