For once the machine opens a door. Nobody has measured how wide.
This month, ministers announced sixty million pounds for new ideas to get disabled people into work. The scheme they already run for that, Access to Work, has been going since 1994. Weeks earlier, the department had published its own research on it, which said in plain terms what thirty years has never resolved: there are still significant gaps in the evidence on whether the scheme even works. So the money for new ideas arrives before anyone has measured the old one. Thirty years. No answer.
That is not a Labour failure. Every government has done the same dance: announce the fund, film the launch, skip the evaluation. Roll it back through the last three and it is the same reflex, hand on heart, every time. A tool is easy to announce. Whether it moves the number is the hard, boring, un-filmable thing, and it is the only thing that counts.
Almost everything we have run this month has been about work going away: the judgment written before the judge reads it, the junior who never gets hired, the safety check that was never tested. This evening is the other side of it, and it is real. For a lot of disabled people, AI is not the thing taking the work. It is the thing that finally lets them do it. A screen that reads itself aloud. A call that captions as you speak. A phone that describes the room in front of you. These are not adverts. They work.
We exist to ask what people do when they do not have to work. For millions of disabled people that question has always had a crueller one in front of it: what do you do when you are not allowed to? The machine can help with that. It can, and I mean it. But a caption is not a job offer, and a fund is not an evaluation, and until someone measures the door, all we have is a very good advert.
The Playbook
Treat Access to Work as a right, not a favour. If you or someone on your team is disabled, it is a government grant for the kit and support a job needs, not a concession. Check what it covers before you assume an app is the answer.
Ask a vendor for the measure, not the demo. When a tool is sold as opening work, the honest question is how many people, into what jobs, and for how long. A demo shows you it functions; only the number shows you it helps.
If you manage people, start with the need, not the device. The gap is widest where the need is hardest to see, in mental health and cognition. The fix is usually a conversation before it is a gadget.
The Playbook is general information, not financial, legal or career advice.
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.
— M.
The counter to the morning's foreclosure story: the one place AI genuinely opens work, and the founding question inverted for disabled people (not "what when you don't have to work" but "what when you're not allowed to"). Whole-bench charge (gate a): the announce-don't-measure reflex across governments, not one party.
Whole-bench competence charge: successive governments announce funds and skip evaluation. Access to Work run since 1994 with no definitive impact evaluation (DWP's own 2018 feasibility study). Levelled across the bench, not one tribe (Note Gate gate a).