Skip to content
2 min read

Sunday Roundup: Biggest Risk Episode Ever, China's AI Inroads, NYC vs AI, Robots and Babies, Cars, etc.

A few things we are thinking about to start the week.


1. AI as the biggest capex risk episode in history

Our longtime friend Jim Chanos gave us a nice shoutout in a recent talk, which was kind of him, in discussing the current episode as one of the biggest and riskiest in Western economic history.

2. China is within the AI margin of error

A new Bloomberg Intelligence report argues that Chinese AI models are within the margin of error, no pun intended, with respect to US frontier models. This is a point we have made many times: models are converging, variance is declining, and benchmarks are largely misleading given their reliance on absolute (not relative) gains.

3. New York City Council is the hero Gotham deserves, but not the one it needs

The city of New York is holding what promises to be a raucous hearing Monday into agents, large language models, and AI risk. It includes "whistleblower" former Anthropic researcher Jacob Coxon, as well as many others. A highlight is that X/AI allegedly ignored the request to attend and apparently had to be subpoenaed.

4. When people compete with robots, they have fewer babies

Like advanced societies needed more reasons to have fewer babies: a new paper presents intriguing data suggesting that competition with robots is causing Chinese families to have fewer kids. The reasons are threefold: training offspring for the race becomes more expensive for parents, their own jobs are less stable, and growing income inequality.