The signal has been flashing red for years and the totals were worrying before covid.

We have had a good metric related to the extent of corridor waits for a long time (the number of 12hr waits). But NHSE resisted publishing the monthly numbers until 2023. Here is what they should have published back to 2010:

It is also deeply misleading to claim that Trump was not prosecuted because of the parts he highlights in the speech. Hd WAS indicted and congress held impeachment hearings which failed largely because of arguments the courts would deal with it so congress didn't need to.

I do like how, in the bit of Prescott's memo where he criticises the BBC for editing together different parts of Trump's speech, his own quote of "what Trump actually said" is... edited together from different parts of his speech.

The USA's healthcare "system" probably does have a big problems with excessive administration costs. But using a bad #dataviz based on bad data does not make a good argument. If anything, it undermines the argument.

The #dataviz problem is to report % growth but not the absolute numbers which is always misleading. But it turns out the underlying numbers are largely bullshit to start with...

They were waiting so they could incorporate important current events like lots of ICE officers on the streets so they could up the scoring system and excitement levels.

The biggest takeaway I have from claims that computers can already do even some forms of general intelligence is that those making the claim do not understand what intelligence is.

It says a lot about the lack of numeracy in the country, and the priorities of the media/commentariat, that there was so much written about the tiny number of pupils spilling over from private to state, and so little about the inevitable large fall in pupil numbers from lower numbers of births.

Fair point. But mostly affect timing by creating a buffer stock when sales are low which shifts time of registration vs delivery/sale. That can't persist for long periods.