4 Comments
User's avatar
Dennis Evanson's avatar

Again, thanks for your impressions and assessments. As always, so helpful, enabling sensible engineered approaches to a sticky problem.

Expand full comment
Mardiros Merdinian's avatar

There are tons of car accidents when drivers blindly trusted GPS routes and ended up in a ditch: https://theweek.com/articles/464674/8-drivers-who-blindly-followed-gps-into-disaster We humans tend to over trust machines than our own judgment. Thanks for the thoughtful piece.

Expand full comment
datasiens's avatar

Brilliant as always. Lets assume we reach AGI then we need to think about how to make the algorithm trust us. What traits in humans will the algorithm find best to establish trust :)

Expand full comment
Dr. Alberto Chierici's avatar

I love the counterintuitive finding about explainability, James.

Expand full comment