Why Europe and the Free World Might Be the Real Winners in the AI War

Stop. Hear me out...

While the US and China are spending trillions of dollars on AI, the EU and developing countries might be getting a free lunch.

In fact, Grok-3 has just been released with open-source weights after xAI spent 20 billion dollars training it.

I will argue that the EU has three core resources that, if appropriately managed, can create astonishing value for the EU and the rest of the world.

1. Open-source AI models

2. Great data and digital infrastructure

3. Close-to-free energy when the sun is shining and the wind is blowing

The open-source models

What we see are increasingly sophisticated open-source models built on the shoulders of giants (through distillation and other forms of reverse engineering)...

These open models (open weights) are now so good that an agnostic model framework can become the de facto standard for nearly all corporate AI strategies worth their salt.

Now the crazy part is that the EU is not really helping here, and the current AI legislation is so outdated and out of touch with reality that it makes me cringe. But still, open-source models are a wonderful gift to humanity and to anyone trying to build something.

Great data and digital infrastructure

A great free model has no value in itself. It is in the application of AI that value is created (the application layer). And here the EU has a massive advantage due to its digital infrastructure and high-quality data, at least in some EU regions like Scandinavia.

I am thinking about health data, population data, government and EU records, historical records, and all the other data that great European culture has created.

To a large extent, this data is already available, and governments should concern themselves with how we safely build on this data while protecting the individual (GDPR is still the best idea the EU has had in a long time).

Good data will help create the best AI doctors and AI engineers, and allow us to run our vast bureaucracy more efficiently, to name just a few ideas.

Cheap energy for cheap training and inference

When the sun shines and/or the wind blows (which happens pretty much all the time in the EU), energy prices can come close to zero.

We need to develop an EU AI infrastructure where token prices are directly based on energy costs, automatically routing token processing to the lowest-cost AI factory.

This would make a pretty good PhD project, so if you are young and clear-headed, please pick up the gauntlet on this one. You might change the world for the better.

Most AI processing can be postponed until the sun is shining and the wind is blowing. A solid open-source framework and some smart EU legislation could make this a reality.

The opportunity

So, to make it clear: the value of AI is not in the technology itself, but in its application. Europe is exceptionally well-positioned to take advantage of AI and to benefit from the trillions of dollars that the US and China are spending.

This is a call to my fellow builders: eat the free lunch and get out there to build something. Like...

• Better crops that do not need pesticides

• Stronger materials for the space elevator

• Abundance for all

We have the opportunity of a lifetime.

./D