[StevenWaterman]

I agree, the first AGI probably will be bodget together with loads of expert input. However, that's not evidence against the bitter lesson.

The first of anything is usually made with the help of experts, but they're quickly overtaken by general methods that lever additional computation

[ksdale]

Sorry, I didn't mean to suggest that the bitter lesson is wrong, just that it's not bitter, it's actually how a whole bunch of stuff progresses.

[StevenWaterman]

Makes sense!