An hour of live television on AI gave viewers one checked number, and it was about water

Tomás thinks this paper has one tune. He was drafted into yesterday's edition by mistake. The Contrarian runs on Fridays, so he runs today, and he has a point.

So Leah has done something different this morning. On Wednesday Sky News gave an hour of live airtime to AI, with viewers' questions and two informed guests. She read all of it, and she has judged it only on what was said. Not on what was left out.

It is still a finding. Work came up at least seven times. Not one number about British jobs was offered in the whole hour. Two answers about AI and jobs, forty seconds apart, pointed in opposite directions, and nobody said so.

And one answer came with its working shown: on data centres and water. It checks out. This paper got that one half wrong in June and put it right a week later. Sky got it right on Wednesday, and we say so.

That is the test I want applied to us too. Where the working is public, the claim can be checked. Where it is not, it is somebody's feeling, however expert the somebody.

One thing I owe readers from last week. On 17 September we asked Otto Car, which Wayve says oversees the drivers in London's self-driving cars, who employs those drivers and on what terms. We gave it until 24 September. Beyond an automated acknowledgement on 21 September, nothing has come. Its accounts for the year to March 2025 were due at Companies House by 31 March 2026 and were still overdue on Thursday evening. Uber, which we asked on 7 September, did not reply by its deadline either. If either answers, we print it.

Leah has the hour. Tomás has his objection. The Playbook is about what to ask when someone tells you what AI is doing to jobs.

A number with its working is worth more than a confident voice. That goes for us as well.

— M.

This note is mine: the view, and the call to run it. It begins as a draft, drawn from work the AI and I have researched and argued out together, the same way every desk in this paper is made, and I answer for every line because I read every line. Those desks run on models built by Anthropic, one of the labs sitting on the very scoreboard we report, so we tell you plainly: we cover this from inside it.