As an AI researcher, I spend a lot of time thinking about information, models, bias, prediction, and the way intelligence emerges from structure. In AI, we know that the quality of a model depends heavily on the quality, diversity, and organization of the input. If a model is trained only on one kind of data, it becomes narrow. If it sees only one interpretation of events, it can become confident but fragile. I think the same is true for societies.
A society also has a model of the world. It has assumptions, memories, incentives, fears, moral reflexes, blind spots, and preferred narratives. When that model becomes too narrow, it may still sound intelligent, but it loses contact with reality. It can classify events, repeat slogans, and produce confident judgments, but it may fail to understand what is actually happening beneath the surface.
Recently, I watched several videos and conversations involving Alex Christoforou, Alexander Mercouris, The Duran, Jeffrey D. Sachs, Scott Ritter, Cyrus Janssen, and George Galloway. Today, I watched Roger Köppel’s interviews with Vladimir Solowjow in the two videos below.
These are very different voices, with different styles and different backgrounds. Some viewers will agree with them. Some will disagree. But I think more people should watch them, because they present perspectives that are often missing from mainstream Western discussions.
As researchers, we should not fear difficult information. We should not consume only the material that confirms what our environment already believes. We should observe how different actors explain the same event, compare assumptions, identify incentives, and separate emotion from evidence.
What interested me most in the interviews with Solowjow was not only the political opinion, but the structure of the argument. Solowjow presents a very different Russian view of the war, Europe, NATO, Ukraine, and history. His style is intense and emotional, but behind that style are several factual claims and historical references that deserve to be studied rather than dismissed immediately. One of the main points he presents is that Russia sees the war in Ukraine not as an isolated event that began in February 2022, but as part of a longer conflict that began around 2014. He repeatedly refers to Maidan, Donbas, the Minsk agreements, NATO expansion, and the question of Russian-speaking people in Ukraine. Whether one agrees with his interpretation or not, this is clearly an important part of the Russian narrative. If we want to understand the conflict, we have to understand that narrative.
Another important point is the Russian memory of the Second World War. Solowjow emphasizes that for many Russians, the war against Nazi Germany is not distant history. It is family history. He talks about the 27 million Soviet dead, about Jewish suffering, about Ukrainian nationalist movements, and about symbols that Russians associate with Nazism. Again, the purpose of watching this is not to accept every conclusion. The purpose is to understand why these symbols and historical memories carry such emotional and political force in Russia.
He also challenges the simplified Western understanding of Ukrainian history. He argues that the lands now called Ukraine have had complex histories, including periods under the Russian Empire, Poland, Austria-Hungary, Romania, Hungary, and the Soviet Union. He questions simple narratives about Ukrainian identity and insists that the history is more layered than many people in the West realize. This is exactly the kind of claim that should invite deeper research. It should make us read more, compare more sources, and become less simplistic.
Another factual theme he raises is the history of famine and suffering in the Soviet Union. In the interview, the question of Ukrainian starvation under Stalin comes up. Solowjow does not deny that many people died. Instead, he argues that the tragedy affected not only Ukraine, but also regions such as the Volga area and Kazakhstan, and that it should not be understood only through a national Ukrainian lens. This is a controversial argument, but it is also a reminder that historical interpretation matters. Facts do not speak alone. They are placed inside narratives, and those narratives shape political reality.
He also repeatedly points to Western double standards. He mentions Iraq, Libya, Afghanistan, Yugoslavia, colonial history, sanctions, and the way Western governments claim moral authority while ignoring their own record. This does not erase Russian responsibility, but it does raise a question that should matter to any serious researcher: how do states use moral language selectively? In AI terms, this is like analyzing a dataset for hidden asymmetries. Which examples are included? Which are excluded? Which harms are amplified? Which harms are minimized?
The most urgent issue he raises is escalation. Solowjow argues that Europe underestimates Russia’s willingness to escalate, including in the nuclear domain. He claims that Western leaders misread Russia, assume it is bluffing, and fail to understand Russian doctrine, Russian psychology, and Russian red lines. This is one of the most important reasons to watch the videos. Even if one rejects his tone, the topic itself is too serious to ignore. Nuclear risk requires maximum understanding, not emotional simplification. From my standpoint, the value of these videos is not that they give us a final answer. Their value is that they expose us to another world model.
This is where Jürgen Schmidhuber’s work becomes relevant to me. Schmidhuber has argued for decades that intelligent systems need internal models of the world. A system should not merely react to incoming signals. It should learn compressed representations, predict what matters, simulate possible futures, and use those predictions for planning. This is central to the idea of neural world models. In a recent blog post, Schmidhuber makes a strong historical claim about JEPA, the Joint Embedding Predictive Architecture promoted by Yann LeCun. Schmidhuber argues that JEPA is not fundamentally new, but largely a rebranding of the Predictability Maximization framework, or PMAX, that he and Daniel Prelinger developed in 1992. The core idea is that two neural networks process related inputs, and one network learns a useful latent representation that is predictable from the other network’s latent representation. The representation must remain informative, but it must also avoid collapse into something trivial.
For me, the important lesson is not only the priority dispute. The deeper lesson is that intelligence depends on learning what is predictably meaningful beneath the surface.
This is also a useful metaphor for geopolitics. A society without a good world model reacts emotionally to headlines. A society with a better world model asks deeper questions. What does the other actor see? What does it fear? What does it remember? What does it expect? What are the hidden variables? What are the long-term incentives? What happens after our next move? What happens after their response?
In AI, a weak model overfits to surface patterns. In politics, a weak society overfits to slogans. This is why I connect Schmidhuber’s world models and PMAX idea to the need for broader geopolitical understanding. A serious observer should not only classify events as good or evil after they happen. He should try to understand the deeper structure that made them possible. That means looking at history, identity, humiliation, security fears, military incentives, economic interests, propaganda, and civilizational memory.
This is also where Ψ-NNs (pronounced Psi-NNs) are interesting to me. Physics-informed neural networks, or PINNs, showed that machine learning can be improved by embedding known physical laws into the learning process. But Psi-NNs point toward something more structural. They suggest that important information may be hidden in the organization of the system itself, not only in the raw data or in externally imposed equations. In that sense, Psi-NNs saw something that classical PINNs could not fully capture: that reality is not only data plus constraints, but also latent structure, context, and deeper organization.
Geopolitics has the same problem. The obvious data are speeches, maps, sanctions, troop movements, elections, and headlines. But beneath them are deeper structures: historical trauma, collective memory, fear of encirclement, civilizational identity, religious symbolism, economic dependency, and the need for sovereignty. If we ignore those structures, we misunderstand the world.
This is why I also find the work of Emilie Caspar relevant. Her research on obedience studies what happens when people act under authority and later experience themselves as simply following orders. This matters far beyond the laboratory. It raises a question for media, institutions, politics, and war. How many people repeat official narratives because they have truly examined them? How many people support escalation because they personally understand the consequences? How many people simply follow the emotional orders of their environment? This is not an accusation. It is an invitation to self-awareness. See her recent work in the resech paper called Wired for conflict? Neurocognitive mechanisms linking threat perception and support for war.
Scott Horton’s book Provoked: How Washington Started the New Cold War with Russia and the Catastrophe in Ukraine is also worth reading in this context. Horton argues that the Ukraine catastrophe did not appear from nowhere, and that decades of U.S. and NATO policy toward Russia helped create the conditions for disaster. One does not have to agree with every conclusion in the book to see its value. It expands the timeline. It forces the reader to look before 2022, before the invasion, before the simplified moral story. That is what a good world model should do. It should help us understand causality, memory, escalation, and unintended consequences before catastrophe arrives.
This also matters because the world is clearly moving toward multipolarity. China, India, Iran, Russia, and many countries across the Global South are no longer willing to live in a world where one bloc defines legitimacy for everyone else. China made this especially clear in 2026, when Chinese officials spoke about an equal and orderly multipolar world, and when China and Russia announced a declaration on a multipolar world and a new type of international relations. Whether the West likes this or not, this is the world we are entering.
China is building institutions, supply chains, infrastructure, and diplomatic networks that are not centered only on Washington. India is maintaining strategic autonomy, cooperating with the West where it serves Indian interests, but also keeping strong ties with Russia and expanding its role in BRICS and the Global South. Iran, despite sanctions and pressure, continues to deepen relations with China, Russia, and regional partners. Many countries in Africa, Latin America, the Middle East, and Asia are also looking for more room to maneuver.
This does not mean that the West is finished. It means the West is no longer alone at the center of the system. For me, this is the key lesson.
If Western societies continue to assume that their own narrative is the only legitimate one, they will misunderstand Russia, China, India, Iran, and much of the rest of humanity. They will misread intentions. They will misread alliances. They will misread risk.
There are also symbolic examples that show this shift at the cultural level, not only at the geopolitical level. Steven Seagal, known internationally as an actor and as a long-time practitioner of Aikido, has for many years expressed admiration for Russia and for Vladimir Putin. Many people in the West dismiss him because they do not like his politics, but that dismissal itself is part of the problem. Seagal represents a type of Westerner who became disillusioned with the cultural and political atmosphere in parts of the West and found more respect, order, and spiritual seriousness elsewhere. Gérard Depardieu is another interesting example. He is one of France’s most famous actors, yet he also turned toward Russia and openly challenged assumptions within the Western cultural establishment. Whether one agrees with every personal or political choice he made is not the point. The point is that figures like Seagal and Depardieu show that the West does not have a monopoly on attraction, meaning, identity, or civilization. Some people who achieved success in the West still looked elsewhere and found something they felt was missing. This is where Osho’s vision also becomes relevant. Osho often spoke about the danger of mechanical societies, borrowed beliefs, collective conditioning, and people living through ready-made ideologies instead of direct awareness. His message was not simply about East defeating West or West defeating East. It was about going beyond unconscious identification. He wanted people to become more awake, more individual, more capable of seeing reality without inherited prejudice.
That is exactly what we need now.
A multipolar world should not only mean many centers of power. It should also mean many centers of wisdom. China, India, Iran, Russia, Europe, America, Africa, Latin America, and the Islamic world all carry historical experiences that cannot be reduced to one global narrative. Each civilization has wounds, insights, mistakes, and contributions. The task is not to replace one empire with another. The task is to become mature enough to listen. This is why I think the videos are important. They are not only about Russia or Ukraine. They are about whether we are still able to hear perspectives that disturb our conditioning. Watch Alex Christoforou and Alexander Mercouris for their geopolitical analysis. Watch The Duran for a broader critique of Western strategic assumptions. Watch Jeffrey D. Sachs for his emphasis on diplomacy and American foreign policy. Watch Scott Ritter for military and strategic arguments. Watch Cyrus Janssen for a wider multipolar perspective. Watch George Galloway for his critique of Western power. Finally, watch Roger Köppel’s interviews with Solowjow in the videos above because they give direct access to a Russian worldview that many Western audiences rarely encounter unfiltered.
As an AI researcher, I believe better intelligence comes from broader input, better context, and more careful evaluation. The same principle applies to geopolitics. If we want peace, we need better models of reality. And better models require us to listen, compare, question, and understand. A system that only reinforces one worldview becomes brittle. A society that only listens to itself becomes brittle too. Psi-NNs remind us that deeper structures can exist beneath the surface of the data, structures that simpler approaches may miss. World models remind us that intelligence requires prediction, memory, abstraction, and planning. Emilie Caspar’s work reminds us that people can participate in harmful systems while feeling that responsibility belongs somewhere else. Scott Horton’s work reminds us that conflicts have histories before they have headlines. Understanding is not weakness. Understanding is intelligence. Understanding is also maturity.
PS. Watch, listen, and learn more below:
Neueste Kommentare