I use first principles thinking to build products that change the world for the better.
My question on this is around management.
Yes awesome that frontier models rock.
But actually managing a nearly infinite group of ai interns {or insert any knowledge worker} is itself a trillion dollar problem. If you've ever tried it, you'll understand degrading returns due to human supervisor capacity.
What's the catch? What innovation did I miss that lets us manage a workforce to advance the scientific frontier? That's what needs to be open sourced yesterday.