Ride the Hardware Lottery!
Issue #3: Why comparing models like GPT-3 with human brain/intelligence is flawed in a unique way.
This post heavily relies on illustrations, which are embedded images. If you’re reading this over email, it might be best to read it directly on substack as some email clients clip long emails and block images.
As you can tell from the previous two posts on Page Street Labs, I have been obsessed with Very Large Parameter (VLP) models lately. I wasn’t alw…