Paula Goldman

Chief ethical and humane use officer at Salesforce; advises the US government on AI policy. Author of Manage the Machine: How to Harness Human-AI Collaboration at Work (not yet ingested). Her team includes roles that "didn't exist a couple of years ago", e.g. responsible AI architect.

Contributions to this vault

• Human at the Helm — her reframing of Human in the Loop: not "AI drafts, human approves", but routing the right decisions to human judgment at the right time. Used as both a system-design principle and a management metaphor.
• "Manage AI as a teammate, but not a human one" — management skills transfer to AI; everyone will manage multiple agents; keep friction so people know they're managing a machine.
• The ~70% rule — "routine to AI, complex to humans" gets you ~70% of the way; the rest is emotions, preferences and sensitive moments (angry → human, embarrassed → AI).
• Bias Toward the Future (Hiring) — pedigree-based hiring blocks internal mobility into roles no one has 10 years' experience in.
• Front-Loading the Brief — via the IKEA couch-in-a-box story: humans ideate first so AI doesn't revert to the mean.
• Entry-level reinvention — Salesforce's counter-cyclical ~1,000 new-grad/intern hire; "managing AI is part of" the new entry-level job. See Deskilling Trap (Juniors).

Reading her

Operator-grade specifics, but not independent: she's a Salesforce executive promoting a book, and Salesforce sells the service agents and HR tooling she describes. Treat her Salesforce examples as a positive-case template with skin in the game (see the tension on Salesforce).

Sources

• How AI Is Changing Talent Not Just Tasks (HBR IdeaCast) — HBR IdeaCast AI-series episode, 2026-09