#466 Who shapes AI: China, USA, Germany, Russia, Ukraine

China with the greatest potential, and still the USA

When it comes to AI, I believe that over the next few years two nations above all will determine the development, and those are China, with the greatest potential, and still the USA. And I believe that as early as 2027 Chinese models will at least reach the performance of American ones and will possibly overtake American models in various areas. And by models I of course also mean agents. The Chinese above all have an open-source approach, while American models are closed according to the classic capitalist principle, and here it’s above all about billions of dollars, while the Chinese are above all betting on market share and on reach. Those are two different approaches.

Germany: top in basic research, far behind in putting it into practice

As far as my nation, Germany, is concerned, I believe that we will keep falling behind. We have never really been a nation when it comes to AI, although you have to draw a clear line here: in basic research we are at the top. German mathematicians are way out in front in the development of AI models, and they work above all at Stanford and in Silicon Valley. But when it comes to putting it into practice, Germany is far behind, because we still rely on fax machines, on fragmented, petty bureaucracy, some of which still comes from the Prussian system, and that holds us back enormously, and I can’t see us untying this knot any time soon. That means Germany will fall further behind. We will play no role when it comes to AI.

Russia will play no role at all in the development of AI

And it’s similar with Russia. There is a fascistoid regime there, which has led to the best scientists in the field of AI leaving, over 100,000, because they have nothing to do with the Putin regime and can’t develop there. And that’s why Russia, too, will play no role at all in the development of AI. Russia will keep stirring up trouble, will keep letting itself be dominated by an increasingly isolated dictator, who at the same time is also a war criminal.

It looks completely different with Ukraine

And it looks completely different with Ukraine. One nation that, alongside China and the USA, will play a major role in AI is Ukraine, because historically – starting back in the Soviet Union – it has produced great top scientists in mathematics and physics, and above all through the IT hub Kharkiv, but also Dnipro and other big cities, it still has a very strong IT community. And unlike Russia, Ukraine is an open country, even if of course it has to deal with corruption. Nevertheless, it is not a dictatorship, it is a democracy, and there, since the invasion, IT and AI specialists have been able to gain experience on the battlefield in the Donbas – above all with drone technology, with mobile data transmission and with data collection, with big data. And that’s why Ukraine, too, will play a major role in the development of AI models in the future, because sadly, through the attack, it is gaining experience every day in real-world application, and other nations will benefit from that. And that’s why Ukraine – and Ukraine’s experience and its specialists – are incredibly valuable for us.

Show fact-check & sources

Fact check & editorial context

The text is Torsten’s personal forecast. The figures and sources here put it in context (as of November 15, 2026).

China and the USA: According to Stanford HAI’s AI Index Report 2025, the gap between the best closed and the best open-weight model on the Chatbot Arena shrank from 8.0% to 1.7% within one year. On MMLU, MATH and HumanEval, the best Chinese models (such as DeepSeek and Qwen) were only 0.3 to 3.7 percentage points behind the best US models at the end of 2024 (Stanford HAI, AI Index Report 2025, PDF). Torsten’s expectation for 2027 cannot be measured yet.
Germany, basic research: German researchers shaped key building blocks of modern AI: Sepp Hochreiter developed the LSTM with Jürgen Schmidhuber (Neural Computation, 1997) and now works at the University of Linz; Robin Rombach, Björn Ommer and their team at LMU Munich laid the groundwork for Stable Diffusion with latent diffusion models (arXiv 2112.10752), and Rombach later founded Black Forest Labs in Freiburg. Sebastian Thrun taught at Stanford and helped build Google’s X research lab. We have no figure for how many German specialists work at Stanford and in Silicon Valley.
Germany, fax machines: According to a representative Bitkom survey (June 2024) of 602 companies with 20 or more employees, 77% still use a fax machine, 25% often or very often, mostly to communicate with public authorities (Bitkom). More on uno.team: #157 · Desktop Faxing – Communication Before the Era of Email, #285 · Power Imbalances and the Fear of Transparency, #53 · Quiet Protest and the Coming Tipping Point.
Russia, brain drain: In December 2022, Russia’s digital minister Maksut Shadayev said that around 100,000 IT specialists, about 10% of the industry, had left the country in the year of the invasion (The Moscow Times). The figure covers IT specialists in general; AI researchers are part of it.
Russia, war crimes: On March 17, 2023, the International Criminal Court issued an arrest warrant for Vladimir Putin for the war crime of unlawful deportation of Ukrainian children (ICC). Yale historian Timothy Snyder describes the regime as fascist (“schizo-fascism”). More on uno.team: #400 · Shameful Hypocrisy: How Germany Betrayed Its Historical Responsibility, #462 · Downward Comparison & Historical Responsibility: Why People Kick Down – and Where Germany Failed.
Ukraine, mathematics and physics: The MESM, the first universally programmable electronic computer in the Soviet Union and continental Europe, was built in Kyiv in 1950/51 under Sergey Lebedev (computer museum); the Glushkov Institute of Cybernetics followed (National Academy of Sciences of Ukraine). In Kharkiv, Nobel laureate Lev Landau worked at the Ukrainian Institute of Physics and Technology; Dnipro, with the Pivdenne design bureau and Pivdenmash, became the center for space and control engineering (The Kyiv Independent).
Ukraine, IT hub Kharkiv: In mid-2021, before the full-scale invasion, Kharkiv had around 45,000 IT specialists and 511 active IT companies (Kharkiv IT Research).
Ukraine, drones and data: The defense cluster Brave1 brings together 2,500 Ukrainian defense tech companies, including over 500 drone manufacturers and more than 200 providers of AI-based solutions (Ministry of Defence of Ukraine). In the Brave1 Dataroom, over 100 companies train AI models on real combat and sensor data (Ukrainska Pravda).
Ukraine, digital government: Over 23 million people use the state app Diia; since 2025 it has included the AI assistant Diia.AI (Ministry of Digital Transformation).