Harvard Professors... Digital Copies
- Project organizers deliberately trained the AI to avoid flattering students
“Al Arabiya.net” – Harvard Business School has launched an entrepreneurial training program that uses digital replicas generated by artificial intelligence of several of its professors, enabling students to converse with the “virtual professor,” practice with them, and receive advice at any time.
The “H.B.S. Foundry” program costs approximately $699, runs for eight weeks, and aims to help participants transform their ideas into startups. It covers everything from idea development and testing, to preparing investor and customer pitches, through to simulating board meetings and sales calls, according to The New York Times.
However, the most striking aspect of the Harvard experience is that students do not interact with a conventional chatbot, but rather with a digital replica that looks, moves, and speaks like a real professor. Among them is Jeff Bossang, co-founder of Flybridge Capital and lecturer at Harvard Business School, who now has a digital “twin” that allows students to conduct video calls with him and practice pitching their ideas.
Moreover, the role of these digital professor replicas extends beyond merely providing information. Harvard designed the AI agents to serve as advisors and coaches for participants. These systems were trained on the work and content of specific professors at Harvard Business School, while drawing on broader knowledge when necessary.
Project organizers deliberately trained the AI to avoid flattering students, making it stricter in evaluating startup ideas. The goal is to subject idea owners to a rigorous test before approaching investors or customers, forcing them to refine and improve their concepts multiple times rather than receiving easy answers or free praise.
This point may be the most indicative of the future of the human-AI relationship. Students can speak with the digital version of their professor at any time, without needing to wait for office hours, schedule a meeting, or accommodate the professor’s timetable.
One participant, a PhD student and startup founder, said he wanted the platform to be available to him at any time he could speak with it, even if it was merely a “replicated version” of the professor.
