January 27, 2026
AI did the homework
Management as AI superpower: Thriving in a world of agentic AI
MBAs speedrun startups with AI—fans say be the boss, skeptics smell hype
TLDR: A professor claims AI tools let exec MBAs spin up startup prototypes in four days, making management skills the new superpower. Commenters split: some cheer the “boss the bots” future, others doubt managers’ chops and question whether the work was even properly reviewed—still, the speed shift is hard to ignore.
A Penn prof says exec MBAs built real-ish startups in just four days using tools like ChatGPT, Claude, and Gemini, arguing the new superpower is management—aka telling AI what to do and when to delegate. The class shaved months off the usual grind, with working prototypes and market research churned out at lightning speed. But the comments are the real lecture. One camp is hyped: cheaper pivots mean trying multiple ideas in parallel without burning cash, and behnamoh drops a meme timeline from “learn to code” to “learn to be CEO,” teasing the era of AI-made companies. Then the skeptics storm in. pglevy throws shade at “management 101,” arguing plenty of managers aren’t great at, well, managing—so don’t assume everyone can suddenly boss around AI agents. The semantics squad shows up too: philipwhiuk wants “AI labs” banned from marketing speak—“it’s product dev, not lab coats.” And the spiciest take comes from ajzushzb, who flat-out doubts the prof even reviewed the students’ work. The vibe: half cheering the “AI intern, human boss” future, half rolling eyes at speedrun startups. Meanwhile, the professor’s delegation formula (time, success odds, review cost) gets meme-ified as a life hack for deciding when to let the bots work.
Key Points
- •Executive MBA students at the University of Pennsylvania built startup prototypes in four days using AI tools.
- •Students used Claude Code and Google Antigravity for prototyping, plus ChatGPT, Claude, and Gemini for planning and analysis.
- •The instructor observed outputs were an order of magnitude further than typical pre-AI semester projects.
- •AI lowered the cost of pivoting, enabling rapid exploration of multiple directions or startups.
- •A delegation model is proposed using Human Baseline Time, Probability of Success, and AI Process Time to decide when to use AI.