Bryan Person

I can't tell you how excited and happy this makes me. Confusing, hard-to-use government software imposes a massive time tax, especially on those who can least afford it (see @AnnieLowrey's work). What if government services were as clear, well-designed, and easy to navigate as Airbnb or Stripe? Hats off to @jgebbia and his team for this exceptionally important work. theatlantic.com/podcasts/202…

I love @benthompson. He is my favorite tech commentator. I listen to @stratechery every day. But his justification for the US Govt seeking to destroy Anthropic is incredibly glib and misguided. AI :: nuclear weapons is sometimes a useful analogy but it's obviously an imperfect one and there are many disanalogies. The notion that Anthropic is developing a power center to rival the US military – which also poses a threat to our country/the world that the US military must neutralize – is wildly out of touch with what's happening, like a 10,000 foot narrative that loses contact with reality at 1,000 feet. It may be clever from a storytelling perspective to draw a connection between Dario's former statements about AI being like nuclear weapons and the current situation, but it's bad analysis. Also, Anthropic is bound by and follows all laws and regulations. In no way has it not "accepted a subservient position" wrt to the US Govt. It is a private company declining (reluctantly!) to enter into a contract with the govt. That is its right. To state the obvious, I'm super disappointed – not because I have some sort of Anthropic bias or political affinity – but because I think the case for invoking supply chain risk against Anthropic is SO OBVIOUSLY wrong and bad. I'm not cancelling Stratechery or anything dramatic, but I hope Ben reconsiders his position. (One of my favorite things about Stratechery is how he changes his mind. Maybe he will not actually change his mind on this – that's fine, that's his prerogative – but I hope he at least gives it deeper thought.)