I am urprised how much money suddenly is available where other future-critical projects are starved of funding while needing just a fraction of that amount, specifically nuclear fusion.
What makes fusion research "future critical"? I honestly don't see any realistic timeline where fusion energy contributes >10% to any nations electricity grid in 2040.
Sure, throwing money at the problem might accelerate things marginally, but you might as well build out renewables, battery storage and invest into grid connectivity, with greater effect (over the next decades), less risk and faster RoI.
It's the same in the UK. Pensioners had their heating allowance removed to plug a £22 bn 'black hole', whilst tonnes of money is pledged to AI. I can imagine the EU is in a similar state.
Removing AI investments will not plug a 22 bn "black hole". This is 200 million europe wide, and the UK is at best 1/6th of Europe if we're very generous.
The EU grants are ridiculously red tapey. Companies have entire departments dedicated to the red tape that European grants require, from the initial participation / auction with hundreds or thousands of pages of documentation, through various stages of thorough checkbox ticking, to the final reception of the project.
Only the usual players with well established "EU Funds" departments can realistically get them, such as Accenture, IBM, Cognizant, Microsoft etc.
I have seen (and worked in) a number of startups that did receive research grants including AI research in Germany. This notion of terrible European bureaucracy is exaggerated. It’s not worse than anywhere else in the world.
I am European and I tend to agree. The way it's currently going, it'll self-select for people who are good at filling out forms, which will either be the giant companies that can afford to hire what are effectively tax-funded civil servants (they happen to be employed by companies, but only exist because of the government and are effectively a tax on business).
Or there'll be a common European subgenre of "startups" that exclusively draw funding from EU/govs because they're great at filling out forms and telling paper pushers what they want to hear. Then they ship some hypercompliant product nobody wants to use.
That is how it feels sometimes. The regulators want to be in control, as if "they are responsible for anything good that comes", while the truth is that they are only able to somewhat restrict or nudge a process, while any strong process will happen regardless of it.
Meanwhile Europe still doesn't have a cloud hyperscaler, so most of these €200B will end up in the coffers of Amazon, Google and Microsoft, the real winners.
I've been thinking recently this is where the EU should start. I don't know if it could be achieved with some top-down initiative, but it would definitely need investment to become one day competitive against the US providers.
Yes, but what kind of AI? LLM chatbots or military robotics? LLMs seem to be drawing most of the money these days and I don't think that should be the focus of the EU. We need advanced drones and anti-drone defense systems, advanced air defense against hypersonic missiles and glided aviation bombs, high precision artillery systems, anti-tank weapons. I don't see LLMs making any breakthroughs in those fields. Maybe transformer neural networks can be tuned to solve some prediction problems in these areas. Maybe some vision models can be tuned to detect and track aerial threats, or movements of troops. My point is, we need investment in AI for defense not in customer support chatbots. Unless it's chatbots that generate smart military moves on the battle field...we don't need them. Let the Americans and the Chinese play with their chit-chatbots. We have bigger fish to fry.
I agree, I think they are trying to cargo-cult Silicon Valley ecosystem without realizing that it will never have a chance in European's clueless bureaucracy, massive social safety nets etc. It doesn't mean nothing can be built, but it won't ever be as effective and bleeding edge for various good and bad reasons and mere fraction of whatever goals aimed for.
And defense spending is a must, how in 2025 they could NOT invest easily 5% of GDP into security is beyond me, thats not a bad requirement from trump actually (if some time is given to reshuffle budgets). There are threats from all sides - one genocidal dictator on our borders repeatedly keeps stating how he will wipe out few hundreds of millions with atomic bombs aimed at our cities. Inner threats with Hungary and Slovakia now.
Then there are global threats, be it China or maybe even US if trade wars will go on 100%, and who knows what Turkey actually wants. Yet Leyden's blind focus on green deal when whole world literally abandoned it, killing our industries while making 0 difference long term is... I don't have a nice word for it. If there is actually some sound logic behind those moves she failed remarkably with communicating them to literally everybody.
They can't invest 5% of GDP in defense because they don't have the money. Germany does not want to raise it's debt ceiling and France is basically broke and hasn't had a functioning government for the last 7 months.
NATO has been treated as a free ride by Europe.
They were supposed to invest a certain % of their GDP each year in their defense spending but because the US was always there as backstop, they basically never did.
Now a lot of countries in the EU have to play catch up and make up the under investment of the last 20 years which will take many years and even possibly decades before it is reversed.
Also Macron is on his way out and the most likely politician to come after him is from the far right. His voice and power is waning and since he has lost the majority in the assembly, he is a lame duck with very little power to change things.
Didn't Cambridge-alumni and London based researcher Sir Demis Hassabis - who co-founded Google DeepMind btw - just win the Nobel Prize in Chemistry?
Comparing the enterprise value of the top 10 companies in Europe versus the U.S provides a poor insight into the actual health of AI research as opposed to capital generation. Both DeepMind and Featurespace are European startups for example.
Maybe they host it on Gaia-X, the failed attempt to compete with AWS and other US cloud providers. Money burns quickly when you give it to EU megacorps that are experienced getting EU subsidies.
In the end, even if there is some moderate success, the EU won't be able to keep the AI researchers because the high tax rates hit hard when you earn top salaries.
Readit News
https://ec.europa.eu/commission/presscorner/detail/en/ip_25_...
• strengthen the EU's generative AI talent pool through education, training, skilling and reskilling activities;
• encouragement of public and private investments in AI start-ups and scale-ups, including through venture capital or equity support
• development and deployment of Common European Data Spaces, made available to the AI community
• GenAI4EU: support the development of novel use cases and emerging applications in Europe's 14 industrial ecosystems
• expansion of computing infrastructure to "100 000 last-generation AI chips"
Overall it sounds very similar to last year's announcement:
https://ec.europa.eu/commission/presscorner/detail/en/ip_24_...
Sure, throwing money at the problem might accelerate things marginally, but you might as well build out renewables, battery storage and invest into grid connectivity, with greater effect (over the next decades), less risk and faster RoI.
There was also the plan go give the Chagos islands to a country that is ~2000km away and pay them billions for the privilege of using them.
There is always money to chase newspaper headlines.
The EU grants are ridiculously red tapey. Companies have entire departments dedicated to the red tape that European grants require, from the initial participation / auction with hundreds or thousands of pages of documentation, through various stages of thorough checkbox ticking, to the final reception of the project.
Only the usual players with well established "EU Funds" departments can realistically get them, such as Accenture, IBM, Cognizant, Microsoft etc.
This is not true, you just have to do a bit more effort to have all your paperwork in order.
Or there'll be a common European subgenre of "startups" that exclusively draw funding from EU/govs because they're great at filling out forms and telling paper pushers what they want to hear. Then they ship some hypercompliant product nobody wants to use.
Deleted Comment
And defense spending is a must, how in 2025 they could NOT invest easily 5% of GDP into security is beyond me, thats not a bad requirement from trump actually (if some time is given to reshuffle budgets). There are threats from all sides - one genocidal dictator on our borders repeatedly keeps stating how he will wipe out few hundreds of millions with atomic bombs aimed at our cities. Inner threats with Hungary and Slovakia now.
Then there are global threats, be it China or maybe even US if trade wars will go on 100%, and who knows what Turkey actually wants. Yet Leyden's blind focus on green deal when whole world literally abandoned it, killing our industries while making 0 difference long term is... I don't have a nice word for it. If there is actually some sound logic behind those moves she failed remarkably with communicating them to literally everybody.
NATO has been treated as a free ride by Europe.
They were supposed to invest a certain % of their GDP each year in their defense spending but because the US was always there as backstop, they basically never did.
Now a lot of countries in the EU have to play catch up and make up the under investment of the last 20 years which will take many years and even possibly decades before it is reversed.
Also Macron is on his way out and the most likely politician to come after him is from the far right. His voice and power is waning and since he has lost the majority in the assembly, he is a lame duck with very little power to change things.
Comparing the enterprise value of the top 10 companies in Europe versus the U.S provides a poor insight into the actual health of AI research as opposed to capital generation. Both DeepMind and Featurespace are European startups for example.
In the end, even if there is some moderate success, the EU won't be able to keep the AI researchers because the high tax rates hit hard when you earn top salaries.