The discussion of Palantir in the NHS is not a question of whether "Palantir are good/bad for taking NHS work," it's actually "should the NHS contract out to (and share sensitive patient data with) a company with a reputation like Palantir?"
The first question makes no sense at all. The latter question, however, is an important one for democratic governments to tackle.
I think the challenging question here is that while palantir obviously have a reputation, they're practically not that different to any of the big consultancies.
Specific to the NHS, Palantir not being a UK company was a big talking point, so yes, that's a tangible difference between Palantir and, say, Capita.
Your (and the sibling) responses also beg the question: must governments contract to big foreign consultancies? It's not illegal to do things in-house if they so choose, you know.
I think by the point you are ready to hire a big foreign consultancy you’ve generally tried in-house and failed.
Of course you can try with a local consultancy, but I wouldn’t know of any, and I assume the reason for choosing one over the other is mostly a matter of reputation.
I think you’re giving the British government far too much credit. Between 2010-2024, the governing party had a religious belief in the power of outsourcing to the point that former ministers (Rory Stewart being one) have spoken out about how many problems they caused by pushing the model to its extremes, irrespective of whether it makes sense or not.
Where it comes to organisational complexity and the barriers it creates, bear in mind that the British state is vastly more centralised than the USA. Fragmentation in the NHS was massively exacerbated by the Lansley Reforms which also forced trusts to outsource a lot of work.
"Your (and the sibling) responses also beg the question: must governments contract to big foreign consultancies? It's not illegal to do things in-house if they so choose, you know."
No, they don't must do this. No, it's obviously not illegal to do things in-house. They choose not to because it's obscenely hard to build what Palantir has already built and to battle test its security anything close to what Palantir has done.
(Disclaimer: I used to work there, so you can go ahead and dismiss my opinion outright, but I am responding directly to what you're saying)
> They choose not to because it's obscenely hard to build what Palantir has already built and to battle test its security anything close to what Palantir has done.
While true, it also doesn’t answer legitimate concerns that the British public had that their medical data was being shared with a foreign entity that had actively participated in foreign government programmes of questionable morality.
The response to that was “all fundamental contractors have done dodgy things.”
To which you have my quoted reply. Which I’m not sure you understood at all, judging by your response.
(See also Microsoft being used to build the French healthcare database.)
The blog's moral stance is that GAFAMs are "neutral" or even "marginally good", because I suppose, they are, among other things, "pro-West".
I don't know which repressive country he "spent a few years in", but I am not sure why he seems more concerned by Russia and China (especially in a country under direct nuclear umbrella) than the risk of parts of the West turning repressive.
And that's not even counting the damage that they might cause outside, like Facebook's complicity in Myanmar's genocide.
That is an important part that a lot of people miss, if you are working with McKinsey you've basically hired some of the worst scum on the earth. Their history on human right's abuses and opioids alone is enough to send them to the shadow realm.
The first question makes no sense at all. The latter question, however, is an important one for democratic governments to tackle.