I don't think any real moat is possible; this is basically maths, and it's not even necessarily such a big advantage to have the best mathematicians anymore. They can try and get political protection but the only real competitive advantage here is to be at the cutting edge.

Show Post

A new post by @samfr.bsky.social is always good to see. A post by Sam prompting a response from @tomski.com is a red letter day. The topic is the reverberations of AI for social change and public services. The implications are huge. samf.substack.com/p/drowning-i... loosemore.com/2026/10/14/t...

Show Post

I keep on coming back to this. There's a reasonable amount of evidence that there'll be a lot of revenue in AI but very little evidence that there'll be access profits. Especially in models. Most AI-driven systems can switch between models, competition remains fierce.

Show Post

Yeah, I mean, electricity one of the great examples of a transformative technology that made very little money.

Show Post

In a way this is the hype scenario: AI crushed a lot of industries, but the AI hyper assume that the AI labs somehow capture that where it seems more likely it just comes out as consumer surplus.

Show Post