Bias Toward the Future (Hiring)

Paula Goldman's phrase (How AI Is Changing Talent Not Just Tasks (HBR IdeaCast)) for the cultural shift internal mobility needs in the AI era: judge candidates on what they can create for roles that don't exist yet, not on pedigree from roles that did.

The argument

HR leaders she interviewed (Seagate, Mastercard, others) report that internal AI talent marketplaces work technically — the tech is "quite mature". Examples: a government-affairs employee takes a small gig on the security team and ends up in a security role; people who take AI workshops and join internal hackathons become forward-deployed engineers.
The blocker they raised isn't demographic bias in AI — it's pedigree bias in managers: hiring for people who "worked at a particular type of company".
For new roles (her example: responsible AI architect), "no one has 10 years of experience with all of these different skills." Big-tech experience is a proxy, but strong candidates will come from unexpected backgrounds; the material question is what they're able to create, and how you assess that.
AI can help assess skills, "but we have to be asking the right questions first and not just defaulting to these kind of shortcuts."

It names the failure mode of reskilling programs: you can train people and surface them via a talent marketplace, and still lose them at the hiring-manager screen.
Operational implication: move role gates from credentials / years in title to demonstrated artefacts (a shipped agent, a hackathon build, a gig-project outcome).
It's the hiring-side complement to Tasks to Responsibilities Shift (the unit of work changes, so the unit of evidence should too) and to Salesforce's skills-mapping platform.

Deskilling Trap (Juniors) — the pipeline problem at the bottom; this is the mobility problem in the middle.
AI Engineer (Role) — a role whose on-ramp is explicitly "learn by building", where artefact-based assessment fits naturally.
The Upskilling Bill — the cost side of the same transition.