Transcription
Here in the UK, Sadiq Khan's decision to block Palantir working with the Met Police has led to attacks on him as valuing ideology over public safety.
Um, however, in spurning reliance on the US tech giant, um, Khan now has significant company. Um, France has this week announced its spy agency will stop working with Palantir, and that's over surveillance fears. Um, Palantir has been used by French spy agency since 2015 and the terrible terrorist attacks, um, they had then. Um, and they've been using Palantir to help sift through huge amounts of data, um, to try and identify, um, terrorist threats.
Um, and as recently as May, France's interior minister, so that's their equivalent of our home secretary, um, told French MPs, um, that alternatives were still not good enough, um, sort of homegrown alternatives were still not good enough to replace Palantir. Um, so Laurent Nunez, uh, their interior minister, um, had said that while sovereign technology should be an aim, it was still not possible to drop American services.
However, more recent events appear to have a shifted opinion in Paris. So, most notably, the announcement last week by Anthropic that Donald Trump was banning them, um, banning Anthropic from providing their most advanced models to European customers or indeed any non-American customers. Um, the news was sort of made public by Anthropic, who made this post on X. So, they said, "The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. Um, the net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected."
This rung alarm bells, um, in Europe because, clearly, we're becoming reliant on technology from a country that's using technology as a geopolitical weapon. Um and that development was explicitly cited by French Prime Minister um Sébastien Lecornu um who announced the new policy on Palantir or not using Palantir anymore for the French spy agency. And so the French Prime Minister said this, "France must have its own tools. We cannot rely on the goodwill of certain partners who, as we have seen in recent days, are capable of cutting off access to the anthropic model."
All sounds very reasonable. There's no pretending though that this will be easy um to support the development of domestic AI. France has pledged an extra 600 million euros in public investment um towards homegrown AI. Um that might sound like a lot of money um but to put that into perspective, the four big US tech giants, so Microsoft, Alphabet, Amazon, and Meta, it's not even including SpaceX, um are planning to spend 700 billion dollars on AI infrastructure in 2026. So 600 million by the French government, 700 billion by the US tech giants. Now that would mean um that the French budget, or at least the budget from the public sector, um would sort of be used at about half a working day in Silicon Valley.
Um Aaron, obviously the logic here is watertight. Like European countries, our country as well, should not be reliant on American tech for sort of crucial national security functions or internal security functions when you have an American government which is willing to use it as a geopolitical weapon and and just switch it off.
>> Yeah.
>> At the same time, those big companies are piling an amount of money into AI development which the Europeans cannot compete with. And so you are in a bind. I mean I I say it's worth it's worth a try. Like we didn't we didn't have Palantir, you know, a few years ago and we still managed just fine. So maybe we can just wait another 5 years until we make our own Palantir, but I don't know, what do you think?
>> So I I don't think we necessarily need Palantir for many things, you know, I think often people have a very expensive technology and they just want to use it. Um I think many issues with regards to public health in this country or policing, actually the the solution isn't technology. All right, the police love to say, "Well, we'll have data recognition systems and pattern recognition systems." Or or maybe just have police officers who are better at solving crime, you know, or who are more motivated for whatever reason, um have better leadership, right? These these things literally don't cost anything.
We saw this by the way with Iran. This is something that's very hard for Westerners to comprehend is that ideology is one example is a is a resource. Doesn't cost anything. Very hard to put a price on the on the on the ideological resilience that the Iranian population had, at least about 10 million plus of them. Um when they were confronted with what America and Israel were doing to them. It's very hard to do that. Uh countries like China, Iran, Soviet Union historically have thought about these things, us less so.
So, do we need Palantir? Maybe, maybe not. Um in terms of the figure we're giving for the United States and the investment they're putting into AI, I mean obviously lots of that's speculative and it's a waste of money. Um you know, should the Europeans be spending billions on this stuff? Possibly. Certainly in terms of things like a replacement for AWS, you know, we don't have really web servers.
>> That's Amazon Web Services.
>> Right? We don't have we there are some cloud There are there are cloud companies in this country as well, but the few significant ones in Europe are in France. You know, Europe needs more of those, right?
>> So that's a data center, right?
>> Yeah.
>> But a data center used for cloud services in as opposed to training AI.
>> I'm giving Yeah, I'm giving one example. We're going to have to spend billions on this stuff. I mean whether or not you think machine learning is good or bad, sovereignty on things like cloud services is going to be really really important. Let's have a conversation about it. I think you're talking about yes, state investment by Paris, but also you do have major companies beginning to emerge like Mistral in France.
>> Yeah, that that that money is not going to go directly on data centers. I suppose that's going to be seed money to try and support a flourishing AI industry in the private sector.
>> So, and also don't forget these major increases to defense budgets. Lots of that's going to go towards AI procurement and if they're not DARPA, which by the way they really are, ideally wouldn't be going to foreign AI companies, be going to domestic ones. So, I'm very realistic about technological autonomy and technological sovereignty. You want You want a sovereign country if your security services use AWS, which I talked about many times. MI5, MI6 use AWS to operate. And if Bezos is told by the American government, kill kill access to that stuff, we're screwed. Same with the F-35s. I'm sure there's kill switch. Sure it's the same with Trident. You need a level of technological autonomy. And it doesn't mean like ultra-nationalism, jingoism, you know, jingoistic pastrami and Michael Walker. Maybe there's shared European capability for some of this stuff. I think there probably has to be, right? But as an anti-imperialist, Michael, I want I want our liberation day day. I want our independence day from the clutches of Trump's America.
>> Andy Thompson with a super chat says, I thought the right-wing media would have been glad to have a large well-armed ship having a pop at a small boat in the channel.