Back to Newsroom

Harvard’s $699 startup bootcamp offers AI avatars of its instructors

By Modelverse Editorial·August 22, 2026·2 min read
Harvard’s $699 startup bootcamp offers AI avatars of its instructors

Harvard Business School’s eight‑week HBS Foundry bootcamp, priced at $699, now incorporates AI avatars generated by HeyGen to give feedback on entrepreneurs’ practice pitches and simulated board meetings. The avatars supplement weekly live instructor sessions, handling the formative critique while humans lead the lectures.

Project director Katharina Rings originally imagined the avatars as simple chatbots, but after a pilot run students asked for more structured guidance. Participants reported appreciating the avatars’ consistent availability, whereas Jeff Bussgang, whose likeness powers one avatar, described his digital twin as “creepy” yet noted that his students enjoy the interaction. A New York Times trial showed the avatar delivering feedback with a noticeably static smile.

  • AI avatars created by HeyGen (no public model architecture disclosed)
  • Integrated into HBS Foundry bootcamp: 8 weeks, $699 fee
  • Provide feedback during practice pitches and board‑meeting simulations
  • Live instructor sessions remain weekly; avatars handle formative critique
  • Student response: positive; instructor response: mixed (creepy but valued)

Why this matters

Fact: The bootcamp uses commercially available avatar technology from HeyGen to scale personalized feedback in a high‑cost executive education program. Inference: By offloading routine critique to AI avatars, the model may reduce instructor workload while preserving perceived instructional quality, suggesting a potential pathway for other costly professional‑training programs to adopt similar human‑AI hybrid feedback loops without altering core curriculum.

Share this article

Found this insightful? Share it with your community on Reddit, X, or copy the link.

ai-newsbrieftechcrunch-ai

Footnotes & Primary References

Related content

Decoding AI's Open-Source Course Maps Three Ways to Run an Agent Loop and the Provider Economics Behind Each

Most teams treat 'which model' as the important decision. The harness engineering literature keeps pointing somewhere else. In LangChain's Terminal-Bench experiment, changing only ...

Read article

Frontier AI labs still won't say how they'd contain a rogue model

A new study finds leading AI labs have few publicly documented plans for containing rogue models, raising questions about preparedness as AI systems increasingly demonstrate unexpe...

Read article

OpenAI says California should strengthen its AI safety bill

OpenAI is calling for California to strengthen SB 53, an AI safety bill that the company previously opposed.

Read article