It's really underrated how the intolerance for friction problem isn't just a problem in learning, it's a problem in getting people who have management authority to respect things like the meaningfulness of safety and security procedures or research integrity.
RE: https://bsky.app/profile/did:plc:oudteec26okqqfisrklsycoq/post/3n2uzwhv26c27
1 Boost
0 Likes
We didn't need the Riemann hypothesis to be proved faster particularly, we don't need autonomous chat bots to be running all our computers more quickly.
But if you do not increase velocity to match all the people who are just doing more with chat bot (and you have to forget the normal verification processes that are typically required for this in order to do more with chat bot) you run the risk of not being in business at all.
Honestly I don't see how this gets fixed until C suite people start facing gruelling punitive outcomes for insisting on bypassing safety/security/privacy processes.
I think this maps on to the AI GRH proof stuff too - a human written proof that's not ready has been published because the author can't afford to wait another year of review and refinement after an AI generated proof was published without the creator doing the work to refine it and make it legible.