Wilbert Copeland

I regret nothing. Bring crackers.

China’s retaliation reminds me of your earlier post about the uncertainty surrounding AI’s maturation in a competitive, multipolar world. Have you previously written about how multipolar (vs. bipolar) AI advancement might uniquely challenge well-being and social cohesion? (Or recommended reading?)

The crux of my interpretation is on the positive-sum aspect. Whether achieved primarily by policy, technology, or magic… I think it’s key to believing that we can do better for the many.

Yea, policy is a major lever. And you are right: it’s complicated to do “well” for so many reasons.

Regulation is key. Didn’t mean to imply otherwise. Maybe swap “let go” with “reconsider and revise”? Not all policy hinders supply, and some hindrance can be good for other reasons.

Thanks!

It focuses on positive-sum actions. Ezra Klein, and others, have been discussing for years. Basic idea: progressives should focus on increasing supply to overcome modern problems of scarcity (eg energy, health, homes). And progressives need to let go of their practices that hinder supply.

Yea, I appreciate this take. A new platform presents new opportunities to grow and thrive. Sometimes the fear of what you might lose can hold you back from all there is to gain.

Plus you get peace by moving past cognitive dissonance.

Hey, I know that guy! Did not realize DSA was on BlueSky. 😊

This could be an opportunity to shift what is being measured. Maybe for the better? Gradually building and refining arguments is important, and good writers do that.

If students have to keep written material in class, then is it bad to let them ponder and refine thought with AI at home?

Why no idea? If we accept evaluation needs to be 100% in class, then why can’t they craft the essay over multiple in-class periods? Perhaps even with assessments on many milestones towards the final essay along the way (e.g., establish thesis, structure, etc).

Also, fallible agents making billions of transactions has potential for significant $$ losses. Will insurance companies move into this space? 🤔

Seems wild how unprepared we are.

Thoughtful! Would AI companies be incentivized against model interpretability to dodge responsibility? Why risk having failures conclusively trace back to their product? Maybe a solution space for AI ethicists?