Recommended reading

Washington watches as West Coast nerds fight for future of AI (2023)

“Young, eccentric techies are fighting for control of the future while operating on a model largely inspired by Mark Zuckerberg’s success turning a web tool for rating his college classmates’ attractiveness into an $800 billion attention-harvesting juggernaut.”

The emerging (Republican) working class majority (2023)

“The choice to finish college and to not finish (or even start) is now the choice that says the most about who you are and what you value in life — between self-actualization in a competitive professional field or an honest day’s work mainly as a way to provide for your family; between acquiring knowledge for its own sake or staying close to the people and places you knew growing up.”

We (Google) have no moat, and neither does OpenAI (2022)

The technology TLDR is that Meta published the full LLaMa model, and improving on LLaMa on a limited budget is outcompeting advances in extremely expensive core model training that Google and OpenAI were hoping would give them a moat.

The policy implication is that regulating AI may go the way of regulating drugs or banning music piracy. It’s astonishing how much policy advocates and pundits simply ignore this, or paint “you can’t put the genie back in the bottle” as a kneejerk capitalist defense, without actually doing the business and science journalism of figuring out if they’re right.