Harvard Business School is now selling a $699 startup bootcamp where AI avatars of its instructors deliver feedback on your practice pitch. Simultaneously, Daron Acemoglu published a sharp piece in Nature arguing we must stop talking about AGI and start building pro-worker AI instead. The HBS Foundry program is a case study in exactly the kind of AI deployment Acemoglu warns against: one that optimizes the institution's margins while cosplaying as pedagogy.
Credential Theater and the Hollow Feedback Loop
The avatar pitch-coach is a perfect specimen of what might be called credential theater. The $699 price point is low enough to feel democratic, high enough to signal legitimacy. But feedback without stakes is just pattern matching. A 2023 paper in Computers and Education by Popenici and Kerr found that AI feedback systems consistently reproduce the biases baked into their training data, flattening the very edge cases that make a pitch compelling. When your critic is a clone, there's no ego to wound, no judgment being risked. The dynamic that makes mentorship valuable, mutual vulnerability, gets surgically removed.
Pro-Worker AI vs. Pro-Institution AI
Acemoglu's Nature argument is precise: the AI industry keeps rebranding the same automation agenda as inevitability. HBS charging for avatar professors is the academic version of the same move. The institution captures the revenue, the instructor's likeness does the labor, and the student is left with a certificate and a chatbot's compliments. Eugenia Kuyda's distinction between home-cooked software and frozen software is instructive here: the avatar professor is frozen software sold as a warm meal. Real pedagogy, like real software, should be shaped by and with the people using it, not licensed out as a simulacrum. The DOJ's scrutiny of a16z's board seat practices this week is a reminder that even in venture, the question of who actually holds the room, and who is just performing presence, is becoming legally consequential.