digital transformation. AI data. innovations and technology.
getty
France’s President, Emmanuel Macron, gets AI. He is saying, in essence, that unless Europe competes at a much higher level in AI, it will end up as a colony of either the U.S. or China.
In the old world of bilateral trade, the doors were open for the U.S. to swamp Europe with Google, Microsoft, Netflix, and Disney. But now Trump has trashed much of that cooperation. Digital services risk being weaponized – or rather, that weaponization has suddenly come into focus.
Any country that wants to be sovereign has to be able to resist software sanctions, restrictions, or weaponization. This is why there is no Google or Facebook in China, and why China has a government policy of phasing out Microsoft as quickly as possible.
The U.S. pulled the rug out from under the world when it locked down Anthropic’s latest AI. That sent a hugely negative message.
In the end, what is more important to a country in the long term: access to artificial intelligence or even something as vital as grain?
Currently, one is an irreplaceable monopoly, while the other is not. One will be vital to the wealth of nations and their economies, while the other is fungible.
There is no second place in intelligence, and there is no natural limit to demand.
That is the key idea.
AI will create artificially more powerful nations that will dominate the also-rans. Spears versus rifles is not a strategy for the weak, nor is begging to the likes of Trump.
Macron has made that plain, and you can be sure the rest of the EU understands it.
So here is the rub.
People are constrained by software.
Software is constrained by hardware.
Hardware is constrained by energy.
This knocks Europe out in a single shot.
Europe has 50% of the electricity for 150% of the people, compared with the U.S. It can’t compete in AI because it doesn’t have the power. The only player in the EU with adequate power – and with the ability to expand it quickly – is France. And what do you know? SoftBank has just pledged billions to build out AI there.
Meanwhile, America has the power and the space to build out… But wait.
China already has 2.5 times the electricity generation and is adding capacity at a staggering pace. It’s building nuclear power at a fraction of the cost of the West. After all, it has been doing so at full scale for decades, while U.S. and EU nuclear expertise has been atrophying. China is already building the energy infrastructure that the U.S. dreams about and Europe is only beginning to recognize it needs.
So catching China looks like an extremely tall order.
The U.S. advantage lies in chips and software, but that moat is eroding quickly. It has funding – perhaps – but the U.S. also has runaway deficits, fractured politics, and resistance to AI from neo-Luddites, none of which are evident or tolerated in China.
The assumption is that America’s advantages are enough to keep it ahead. Yet technology moves far faster than infrastructure. While a technology moat can disappear in months, the energy gap may never be closed in a political system where a rare bird nesting can indefinitely halt the construction of a power station.
Musk talks about robots. China sells them – today. And they don’t stagger around an office; they perform kung fu.
China can and will close the U.S.’s software and hardware moat.
Can the U.S. close China’s energy moat?
Macron said Europe has committed about $600 million to Mistral AI, while the U.S. has invested roughly $600 billion in AI. He believes it is remarkable that an investment amounting to just 0.1% of the U.S. total has achieved so much. But he also recognizes that it is not sustainable for Europe’s sovereignty to remain at the back of the AI peloton.
I can develop this analysis in two directions from here:
- America cannot build enough energy quickly enough and is finished, or
- There will be a truly titanic build-out of energy infrastructure in the U.S. and Europe to catch up with China.
Version one is much more commercially attractive for me – people love doom – but I think we are going to see a massive pivot toward energy.
It is coming very soon, and if that effort sputters, then we can all choose option one.
However, if it doesn’t, the economic consequences will be enormous: rapid growth, high inflation, and a secular increase in economic and political volatility for years to come.
Energy is the key to our future, and it can be bright if we choose AI over NS.
NS?
Natural stupidity.
Sadly, there is no shortage of that.