gen‑ai.news
← Back
Multimodal

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

Harvard Business School has launched a program called HBS Foundry, a $699 online bootcamp aimed at early-stage founders, with a feature that sets it apart from typical online courses: AI avatars built to resemble and respond like the program's human instructors. These avatars are used specifically in interactive exercises - practice pitch sessions and simulated board meetings - where participants receive real-time feedback as though they were in the room with an actual faculty member.

The use of AI avatars in this context reflects a broader trend of institutions experimenting with generative video and conversational AI to scale access to expert guidance. Rather than pre-recorded video alone, the avatars are designed to be responsive, giving the impression of a live exchange. The $699 price point appears to be a deliberate effort to position the program as accessible compared to traditional executive education, which can cost many times that amount for even a few days on campus.

From a technical standpoint, programs like this typically rely on a combination of video synthesis tools - capable of animating a realistic likeness - and large language models trained or fine-tuned on an instructor's known frameworks, vocabulary, and teaching style. The quality of feedback such a system can provide depends heavily on how well those underlying models capture the depth and adaptability of the real person. A scripted pitch scenario is a relatively controlled environment, which makes it a reasonable starting point for this kind of deployment.

The educational implications are worth watching. Proponents argue that AI avatars allow learners to practice high-stakes conversations repeatedly and without the scheduling friction of real faculty access. Skeptics will note that a simulated board meeting can only approximate the unpredictability of real investor dynamics. What HBS Foundry ultimately tests is whether AI-mediated feedback, even when it carries the face and name of a credible instructor, can meaningfully move the needle for founders preparing to go out and raise money or run a company.

Enjoy this story? Get the next one in your inbox.

Twice a week: the most important stories in generative image and video AI, distilled into a 2-minute read.

Free. Unsubscribe any time. No spam, ever.

Your next read

New benchmark confirms AI models still perform poorly at visual perception
Multimodal

New benchmark confirms AI models still perform poorly at visual perception

A new benchmark from Moonshot AI isolates visual perception from logical reasoning in multimodal models, and the results are sobering. No tested frontier model clears 60 percent accuracy, with GPT-4o leading by only a narrow margin. The findings suggest that many errors previously attributed to faulty reasoning may actually originate much earlier, at the point of reading the image itself.

You can now turn off Google Gemini’s visible watermarks
Multimodal

You can now turn off Google Gemini’s visible watermarks

Google has added a toggle in Gemini and its AI video tool Flow that lets users remove the visible "sparkle" watermark from AI-generated images, videos, and music. Even with the visible mark turned off, content will still carry invisible SynthID watermarks and C2PA metadata. The change affects content produced by Google's Nano Banana and Omni models.

Twitch streamers can now opt out from training Amazon’s AI
Multimodal

Twitch streamers can now opt out from training Amazon’s AI

Twitch has introduced an opt-out setting that lets streamers prevent their content - including streams, VODs, clips, chat logs, and channel images - from being used to train Amazon's generative AI models. The control applies to future training only and covers AI systems designed to generate or synthesize text, audio, images, or video. Non-generative AI features such as captions and safety tools are unaffected by the setting.