Conversation

The other day I described my concept of "feats of skill" (e.g. Everything Everywhere All At Once) & "feats of budget" (e.g. Avatar franchise).

I'm much more interested in feats of skill, which explains why I can't bring myself to care about GPTs (which is all based on "how much data can we feed into these neuralnets?"). No matter how much it feels like I'm expected to!

New policy: You try to evangalize GPT to me, I'll evangalize Haskell to you. That actually addresses the bottlenecks I face!