It’s a bit of a grim take.

For example

> The poor could exercise their lawful rights as successfully as the sharp-elbowed middle classes do today. Yet it threatens to become a tragedy of the commons. If the state is overwhelmed and cannot function, everyone loses. When 20th-century governments created broad rights, they had the noble ideal of making citizenship meaningful. The public would be heard in consultations, get information under transparency laws and win redress for maladministration from a panoply of ombudsmen, tribunals, commissioners and judges. But these analogue systems assumed that few people would have the time or temperament to pursue their rights to the bitter end; and that of those who did, few would have the money to pay for a lawyer.

Seems like the implicit admission is “the system was staked against the poor, now that is not anymore we call it broken”.

Maybe I’m wrong at assuming this is also skewed towards people with less financial know-how, or easy access to an accountant, but here it is:

> Britons under-claim the benefits to which they are legally entitled by as much as £20bn ($27bn), or 0.7% of GDP, a year; if everyone gets their dues, the public finances will wilt.

Even more so:

> Money should be part of the solution. Britain could charge a fee for freedom-of-information requests, as other countries do. So could employment tribunals, which should also be more able to award costs against opportunistic claims.

I am not familiar with the inner workings of British administration but the tone of the article makes me feel AI as a democratising force is depicted as dangerous because it works at achieving a noble goal: equality in opportunity of interacting with state.

And it turns outright dystopian when

> The most important task is to act faster and more radically on remaking the state with AI.

Advocating for a fast and radical adoption in embedding AI in taking decisions over citizen rights is the opposite of what I would consider sound judgement