"What are the chances for Germany and Europe in terms of AI? Amy Webb is one of the best-known futurologists in the world. In an interview, she reveals why she believes that Germany has two problems.
Have the USA and China already won the competition for artificial intelligence? The international market share of American internet or platform companies is currently around 70 to 80 percent. Europe reaches 1.5 to a maximum of 3 percent. And we see something similar in the IT competition, where China seems to be catching up a bit. Where is it headed?
Amy, what do you think? Is AI currently a game dominated by America?
No, the competition is far from over. An incredible amount of work and money is being invested in the United States and China. There is also a very high level of commitment in the United Arab Emirates and some other places in the world. But the story is much bigger.
Claude or ChatGPT and the other systems do not work without hyperscalers. That means they need the cloud. And they require insane computing resources. That means today: you need Nvidia. But Nvidia also has competitors, that's important. The question of the future of AI is really about the future of a multifaceted ecosystem that has many relationships and many inter-dependencies.
And where does that leave Europe?
In France, you have Mistral, which I think is one of the leading companies right now. Open source and completely open AI systems - that offers enormous flexibility and opportunities. And the concept is being adopted by other companies right now.
On the other hand, there are security concerns. And for me, it's ironic that this is happening in Europe right now. It's perhaps an unintended consequence of some regulations. All these regulations that were actually meant to fight Big Tech led to Mistral, which on the one hand is great because it's open and not tied to the other big companies.
But: it's absolutely not safe. I mean, they're releasing new versions of Mistral on Bittorrent, right? There's no documentation to go with it.
You mentioned Saudi Arabia and the Emirates. One of Europe's leading AI scientists has been there for more than two years now to help build an ecosystem. What are the chances of success?
The Emirates and the Saudis are competing with other countries for the future of technology. What I saw in the Emirates was quite "pretty" and open. Very keen to be completely neutral. The Emirates have Chinese servers and Chinese language models that run things, but they also have American and European ones.
It's the same in many fields, whether we're talking about artificial intelligence or synthetic biology. I was in a company. There are American sequencers and European and Chinese ones, and that's a smart strategy as the world is being reordered politically. I think trying to establish yourself as a neutral area is a smart move. But to be fair, the United States has far more researchers and publications. It's going to be hard for others to catch up.
The Chinese have invested a lot in the Emirates and the US. Is the country catching up quickly?
It depends who you talk to. Some of the people I've talked to about the future of AI and China in particular will point to the sheer volume of patents filed by China. It's going through the roof.
On the other hand, what's the result? A patent is not a product. And: They are patents that are recognized in China but not necessarily valid in other places. But if you think about the number of academics in China, you see a factor of ten compared to the US.
Ultimately, I don't think this is a race where someone crosses a finish line first. Even if you think about other fields like quantum computing. It's less of a race in terms of technical development. But there is certainly a race for market share and capital. We're seeing that race 100 percent right now.
So it's a global race. Every continent is involved. But let's focus a little on Europe. What is your opinion on the state of play in Europe? Where are we in terms of AI?
I think when you talk about the European economy or politics in Europe, you can't look at an entire continent when it comes to technology and specifically artificial intelligence. I think it makes sense to talk about Europe in a differentiated way because the performance is predominantly in two places. Well - two and a half.
That's Germany and France. Before the conflict, there was Ukraine, and now Poland is kind of stepping in. I don't see the same movement in Portugal, Spain, Italy or Belgium. It just doesn't happen there.
I think that's an important difference because in Germany, AI and the German economy go hand in hand. The backbone of the German economy is manufacturing, auto and pharmaceutical. They are areas where AI will have the biggest impact at the beginning. So that makes sense.
But when I look at France, you know, France's biggest industries are luxury goods. Then what else? We have to talk about the future of Germany if we want to talk about the future of AI in Europe - even if Mistral comes from France.
Okay, let's talk about Germany.
The problem in Germany is that the country doesn't take enough strategic risks. There are great research teams at some of the world's best universities. They have good companies in industries like auto, pharmaceutical and so on. But these companies are not ready for AI.
There is something like a "loose connection": I don't see these companies investing (even) mentally to somehow create new growth opportunities in the future. Yet Germany should be so far ahead of China in cars.
If you went back ten years and asked the average person whether China would replace Germany as the global exporter of cars. Everyone would have said, what the hell are you talking about? Right?
But that's where we are now: cars and AI are very closely linked. AI supports manufacturing plants, sales, logistics, supply chain, development, testing and building design. Not to mention the quality experience of sitting in a Chinese car.
The same goes for the pharmaceutical industry. People say, look at the Covid vaccine. A German company was able to bring a product design to market that used artificial intelligence. But that's just one example.
I look at all the other pharmaceutical companies that just don't do that. They don't invest enough in these new areas of artificial intelligence and new drug discovery. Maybe their venture capital firms are invested. But, quite honestly, these are huge companies that should now start to get active here. And I don't see that happening at the moment.
If you look at Volkswagen, they failed with Cariad and trying to develop their own operating system for a car. Are they too big and not flexible enough to change the house to a software-driven company?
Germany has two problems. A culture problem and a leadership problem. And these two are related. The culture problem is that nobody wants to be wrong. There is a pathological fear of making mistakes. This leads to me getting a "no" all the time first and then a "maybe" or "we'll try".
But this is also related to leadership. I see that in companies outside Germany, KPIs (key performance indicators) and performance are more closely linked to innovation. Here is an example:
I can't say who it is, but it's a German company. There is a board of directors and a supervisory board. But they are just paralyzed. Changes should be made. The company is well capitalized, has a vision, and they have a lot of money in the bank. So there shouldn't be any bottlenecks to at least think about maybe trying something different at some point. But: The board just said no, they want to focus on the next few quarters. Now is not a good time to try something new.
But where does this reluctance come from?
I don't know exactly. But maybe there is no incentive to be different, to do it this different way.
The opposite of that is a company in Germany whose board I sat down with recently. This company is going to completely change its industry. It's a big, global company. What they did really intelligently is that they built a prototype. Quietly, to test a theory, to see if something could be different than what they've done so far.
And theoretically it works. They're going to try it out a bit soon to see if it works. They'll learn as they go. And then if it works, they'll scale it up. But that came from their CEO, who said, I think there is a chance. It might not work out the first time. But that doesn't matter: you have to have a vision. And maybe that's what's missing. This CEO I'm talking about has a crystal clear idea of what the future of this industry and this company will look like. Can it fail? Sure. But then you have learned something.
But I often observe with German companies that perfection is the goal. Perfection or nothing.
Do you have a recommendation for politics? What about European regulation?
Regulation is difficult. I understand the desire to have guard rails to secure the future. On the other hand, you have to be able to enforce regulations. And if the regulations are specific, it is easier to enforce them.
On the other hand, these regulations become outdated because regulation is by nature a snapshot. And it is difficult to predict what will happen. Here is an example that Germany and Europe could perhaps follow. Abu Dhabi has a beta testing policy. The first time they applied this was to drones. They said: let's think up scenarios for the future. Let's get started. So they did sandboxing and tested what the guidelines for this technology could look like. After that they think about regulation.
Finally, what else is fascinating to you right now besides AI?
Quantum computing. I think sometime in the next three to five years we'll see some kind of hybrid quantum in a high performance computing center.
And if you combine that with AI?
Sure. You'll actually need that, because for AI to evolve, we'll just run out of computing power at some point. Not to mention it's expensive and environmentally damaging.
What is your expectation of an AGI? So a General Artificial Intelligence. When is it coming?
I would say AGI is here. It just hasn't shown up in the way everyone expected. Part of the problem is that there's no common definition for it. Open AI has taken the liberty of defining itself as an AGI company. I would argue that AGI is already here. It's better to acknowledge where we are and where we're going than to look at the forecast of someone predicting when something will happen. Right?" [1]
OK, France is stuck with stupid luxury goods that do not fit in the AI era. Germany has an industry. Under stupid government of Germany's leader Scholz, Germany is de-industrializing rapidly. France is pulling Germany out of its biggest market - China. America with subsidies is sucking out investment and talent from Germany. Scholz allowed foreigners to cut German industry from cheap Russian energy making it impossible to compete on price. German industry is dying. AI will help German industry the same way a dead person is helped by a hot bath.
1. Der KI-Wettbewerb ist noch nicht entschieden. Frankfurter Allgemeine Zeitung (online) Frankfurter Allgemeine Zeitung GmbH. Aug 28, 2024. Von Holger Schmidt und Johannes Winkelhage
Komentarų nėra:
Rašyti komentarą