Welcome to AI in the Real World. I'm Joanne Chen, a General Partner at Foundation Capital. On this show, I speak with leading AI researchers about how cutting-edge models are being applied in enterprises today.
Joining me today is Eugene Cheah, the founder of Recursal AI. Recursal is working to commercialize RWKV, a new AI architecture that Eugene played a key role in developing.
Eugene believes RWKV can address some of the main limitations of transformers - the technology behind models like GPT and Claude that have fueled the recent AI boom. By combining elements from both transformers and RNNs, RWKV aims to make major improvements in how efficient, scalable, and accessible AI can be.
In our discussion, Eugene takes us through his journey, from the early days of building RWKV's open-source community, to launching Recursal to commercialize it. He breaks down how RWKV works, what sets it apart from transformer models, its potential benefits, and its most promising use cases.
Read the notes from our conversation here: https://foundationcapital.com/whats-next-after-transformers/