[paulsutter]

It's much more important for you to help your friend come up with a real world problem that he can solve (whether or not it uses his technology), and for him to start solving that real world problem for customers. Dont bother thinking about VCs until he has a proven scalable business opportunity.

I'd be happy to help brainstorm that (sutter at gmail)

Arcane machine learning challenges, by themselves, aren't very interesting to VCs or customers. But solutions to real world problems are.