← Back to context

Comment by foota

1 day ago

From their github: "The average result used the equivalent compute of roughly three hours of ChatGPT Pro thinking." That's pretty crazy.

This is what really made me think to my self, "holy shit". I can't believe not more people are noticing and talking about this. Unless perhaps they didn't actually read the README, and are just talking about what they heard from someone who also didn't read it?