Notes from Molly’s presentation:
- Radical exclusively invests in early stage AI
- Firm’s origins are in Toronto, the location of many key players, institutions, and events in the AI histroy, including Geoff Hinton of the University of Toronto, winner of the Turing competition.
- They see themselves at the center of the center of the AI research and talent ecosystems.
- They look for researchers coming out of a lab.
- Radical is on their second fund of $500M. They invest across the AI stack.
- They’ve helped portfolio companies ike Cohere and Reka transition from research to commercial.
- Velocity is their accelerator program.
- If you’ve ever tried to hire AI or ML talent in the past, you know how hard that is.
- Radical helps with AI compute as well as unit economics. They know startups need talent and compute.
- They invest across the AI spectrum, from application areas to digital infrastructure.
- 2023 is the year of computer. Computer power is increasingly available and affordable. Stable Diffusion only took about $600k to train.
- 2024 may be the year of the AI application.
- There’s a lot of lobbying and regulation in the US and beyond around foundational models, especially as they apply to finance and healthcare.
- Macrofund’s changing thesis is that AI will be software, and software will be trained by AI.
- Radical likes technical moats in the application layer. There was some audience Q&A around open source and the potential of miniaturized models that run on mobile.
- Future applications will need security, privacy, and ease of use.
- Q&A around watermarking, especially with AI created synthetic media. This is a problem a lot of people are working on. Molly does not currently see anyone using blockchain technology to watermark synthetic media.
- They’re looking for the “Attention is All You Need” paper of today.
Learn more about AI, VC, and startups, at inside.com.
