UCLA published findings on how various Language Models performed on analytical reasoning questions against an average college applicant. GPT-3 is a clear winner against the average applicant! 🤖 https://i.imgur.com/aYNFsQK.jpg

🇧🇷🇺🇸-  Book: Making Things Think: https://holloway.com/mtt. Investor in Wander, Carry, Footprint, Merkle Manufactory (Farcaster), Dynamic, Paragraph.

and yet it's still bad at counting... why?

The only problem is externality pricing.

Building https://negationgame.com

I asked it to type the word "yes" 150 times and it refused on grounds that it's an AI. But when I asked it to write a program that printed 1 to 100 it complied.

It's like it was annoyed by the simple task

In my experience, it is pretty good at counting... https://i.imgur.com/6IhYYKZ.jpg