The state of the near future is changing. Like an office building without employees, it functions formally, while real power shifts elsewhere
Artificial intelligence is no longer just a question of who has stronger machines and who can afford them. It is beginning to enter the very architecture of the state, the way decisions are made, implemented, and justified.
The American “Genesis Mission” (the name itself) perfectly describes this moment. It suggests the beginning of something that can no longer easily be reversed. Genesis is not a reform, not a law, and not a strategy in the classical sense. It is an attempt to create a new foundation, a new operational layer of the state—one that is not visible in constitutions and election laws, but which in practice determines how the system functions. While the public continues to focus on political conflicts, elections, and ideological divisions, beneath that level an infrastructure is emerging that does not debate, negotiate, or delay. It calculates, compares, optimizes, and moves toward managing the system.
The state as we historically know it was slow precisely because it had to be slow. Decisions passed through procedures, discussions, committees, corrections, and resistance. That process was imperfect, often inefficient, but it had one crucial role: it gave the decision political meaning and social legitimacy. Artificial intelligence introduces an entirely different logic. It does not necessarily ask what is just, popular, or acceptable, but what is statistically more efficient, safer, or more stable in a given model. When such logic begins to be embedded in administration, security, science, and governance, the state no longer decides—it increasingly just confirms what has already been calculated.
In this context, the “Genesis Mission” is not just an American technological project or Trump’s political gesture. It is a symptom of a broader change in which power shifts from visible institutions to invisible systems. Parliament, government, and ministries do not disappear, but they slowly lose the role where decisions are truly shaped. They increasingly become, let’s put it that way, a “public interface,” while real processes take place on platforms that connect data, models, and infrastructure into a single whole. If this trend continues, the state as we know it could remain formally intact but functionally hollowed out, like a building whose walls are there, but life and work have moved elsewhere.
This text does not deal with whether artificial intelligence is good or bad, nor does it attempt to predict apocalypse or a technological paradise. Instead, it tries to answer a simpler but perhaps more uncomfortable question: What happens to politics, democracy, and sovereignty at the moment when machines begin to make decisions faster and more consistently than people, and politicians gradually turn into their translators and ambassadors. Artificial intelligence is now becoming an emerging form of governance, and that is the true genesis we are witnessing.
The Genesis Mission as an Institutional Precedent
When a new initiative is announced in Washington, especially in the field of technology, it is most often a combination of political marketing and real intentions. The “Genesis Mission” is interesting because it does not stop at phrases but immediately enters what in practice gives states power: data, infrastructure, and the ability to connect it all into a system that works every day, all day. The published executive order openly talks about an integrated platform that is supposed to unify federal scientific data, supercomputers from national laboratories, and a secure computing environment in the cloud, along with models, tools, and even experimental and production capacities for autonomous and AI-assisted processes.
On paper, it looks like accelerating science, like an attempt to shorten years into months. In practice, it means the following: Instead of each laboratory, agency, or research group working in its own “silo,” with its own rules and limitations, Genesis envisions a common system where data on materials, energy processes, or biological structures can be analyzed on the same computing resources and the same models. For example, simulations for developing new nuclear fuels, which until now were conducted separately and lasted months, are supposed to take place within a unified environment, with AI models that simultaneously test thousands of variants and suggest the most promising directions for further research.
The same applies to energy. Genesis explicitly mentions the application of artificial intelligence in optimizing electrical grids and in developing fission and fusion technologies. This means that data from real grids, production facilities, and experimental reactors can be unified into models that assess system stability in real time, predict failures, and propose operational decisions. In other words, AI is not used only for analyzing past data but becomes a tool for everyday management of critical infrastructure.
The order very clearly states that the platform must enable training and use of large models, development of specialized “agent” systems, and autonomous experimentation. In practice, this means robotic laboratories where AI not only analyzes results but also proposes the next experiment (a very important point), changes parameters, and accelerates the entire research cycle. Such systems already exist in limited form, but Genesis tries to turn them into a standard state tool, especially in areas declared strategic, such as biotechnology, advanced materials, and semiconductors.
The pace prescribed is particularly telling. The Department of Energy is tasked with creating a complete inventory of available computing capacities, networks, and data in a short time, then defining initial datasets and models, involving resources from other agencies and private partners, and demonstrating a functional platform example on a concrete national challenge within just a few months.
This is precisely why Genesis is an institutional precedent. It is not about buying software or funding individual research but about attempting to unify in one architecture what has until now been deliberately separated. The dispersion of data, computing resources, and jurisdictions slowed processes but also prevented the concentration of power. Genesis goes the opposite way and builds a coherent system that can act quickly and coordinately.
It is no coincidence that the project leader is the Department of Energy. It is an institution that in the American system has for decades connected science, industry, and national security, including oversight of the nuclear arsenal. Entrusting Genesis to that institution clearly shows that artificial intelligence is viewed as a strategic tool that must be deeply integrated into the very structure of the state. In that sense, Genesis is the first serious attempt to build a state platform that acts as a nervous system of a new type of governance.
When the State Stops Deciding and Starts Optimizing
In the classical state, politics has its own theology. It is built through debate, through compromises, through slowness, through procedural rituals that are often tedious but serve to give the decision legitimacy. In AI logic, there is none of that. It is cold, accounting-like, and at best utilitarian. It does not ask what we want as a society but what is more likely, faster, more stable, or cheaper in a given model. Here we are talking about a large, state-level system, not about AI models “for the masses” like ChatGPT or Grok, which are “programmed” (depending on which) to be caring, cautious, even friendly, almost human.
In this state version, AI is something very different and is not seen as a tool for one sector but as a mechanism for managing large systems. In the most optimistic interpretation, it means accelerating innovation and solving problems like grid stability, more efficient production, or developing new materials. In the political interpretation, it means shifting more and more segments of the state from the sphere of decision-making to the sphere of optimization—in other words, as if society is a machine that needs to be “properly tuned.”
In Genesis itself, this optimization impulse is very visible. It mentions automation of experiment design, acceleration of simulations, and creation of predictive models for diverse domains, from biological processes to plasma physics in fusion.
Of course, knowledge has always been the foundation of state power. States differed not only by army but by the ability to organize industry, science, and administration. In the AI era, that ability is increasingly measured by who can faster turn data into recommendations and then recommendations into operational decisions. This is where the thin but crucial change happens: AI is no longer just support for the human who makes the decision but becomes the environment in which the decision is already shaped before the human appears.
And that is why, obviously, it is important that Genesis does not choose “neutral” areas. The order explicitly says to focus on at least twenty “national challenges,” and those challenges must cover domains like advanced manufacturing, biotechnology, critical materials, nuclear fission and fusion, quantum science, and semiconductors. You will quickly notice that these are not topics of scientific curiosity but the foundation of industrial and military power.
We mentioned it, but let’s mention it again—the most important thing is to understand that optimization is not neutral. It always assumes a goal. When an algorithm optimizes, it optimizes according to given criteria. And criteria are not set by mathematics but by politics. The difference will be that in such systems, politics hides. It is no longer public debate. Is that important? It should be, but everything currently looks like things are moving faster than the public can follow. As we said, this is a process that will not reverse, so once it “settles,” that will be it.
Geopolitical Mirror and Three Models of the AI State
Great powers do not compete only in technology but in what kind of state they want to build around the technology. And three different models are already visible, each with its advantages and internal contradictions.
The Chinese model is most often wrongly portrayed as a simple caricature of surveillance, although reality is much more complex. China, back in 2017, set the goal in strategic documents to become the leading global AI innovation center by 2030, and that goal did not remain in the realm of rhetoric. It is an industrial policy covering the entire chain, from chips to application, and at the same time a way of managing society in which technology is seen as a tool for coordinating large systems. Recent examples of AI application in energy show this logic in practice: AI systems are used to align production and consumption in grids with a high share of renewables, for virtual power plants coordinating different sources, and for industrial projects trying to optimize processes related to hydrogen and chemical production. In other words, AI is not treated only as a “digital sector” but as a management layer for industry and infrastructure, in line with the tradition of planned direction of key branches.
The European model is seemingly opposite. Instead of a platform, Europe first built regulation. The AI Act entered into force in August 2024, and the first bans on certain “unacceptable” practices began to apply in February 2025, with further phases relating to general-purpose models and high-risk systems. In the European approach, the state primarily appears as a regulator and guardian of procedures, and only then as a builder of infrastructure. However, what is becoming increasingly visible is that this model enters a crisis as soon as it collides with economic reality and lobbying. During 2025, the Commission publicly rejected the idea of delaying implementation, only for information to later emerge about proposals that would postpone part of the stricter rules, especially for high-risk systems, until 2027. In translation, Europe tries to be a “sovereign regulator,” but without its own platform and without its own global champions on the scale of American and Chinese ones, so every serious regulation immediately turns into negotiations about competitiveness and who will bear the cost. It is not hard to predict which of this “big three” will end up last.
The American model, at least in the form of Genesis, tries to be a third way, although in essence it is very recognizable. The state does not build plan-wise like China, nor does it primarily define itself by regulation like the EU, but uses executive power to direct and accelerate development, and the private sector to operationally execute it. In practice, this means creating a platform under state political patronage but with private components that carry the largest part of the technology, cloud infrastructure, and often the models themselves. The official language emphasizes dominance, security, and reducing dependence on “foreign adversaries,” but the mechanism to achieve it is public-private symbiosis in which private actors become an indispensable part of state function.
Three models, three different policies, but the common consequence is the same: AI is pushed into the center of state action, and with it a new kind of power that is hard to control with classical democratic tools. For smaller states and the periphery, the difference is often only in whose architecture they will connect to, and under what conditions, because they generally lack their own model for building a platform.
The State as a Shell and Politicians as AI Ambassadors
Once the idea of the state as a platform is accepted, futurism almost becomes a continuation of analysis. Not because parliaments and ministries will disappear tomorrow, but because something more banal and dangerous can happen: institutions remain the same, but their content changes. In that scenario, the state does not collapse—it is “emptied” from within, and real management shifts to systems that produce recommendations, scenarios, and “best options” faster than politics can discuss them.
Genesis is a good example of how that process could look because it already emphasizes in the official description systems that explore the solution space, evaluate outcomes, and automate workflows. In a laboratory, it looks like accelerating science, like optimizing experiments. In administration, it would look the same—as accelerating the state, as optimizing politics. Once there is a mechanism that can produce hundreds of regulatory variants in a few hours, assess their costs, calculate political risk, and predict consequences, human debate just slows things down.
In that picture, politicians do not disappear—they also change function. They become ambassadors, translators, spokespeople. Their role is to explain to the public what “we must” do, often with the argument that the numbers are clear and the system has shown the optimal path. Some countries, we could say, already have such a model! Take marginal and politically less important EU members as an example. They have governments that just “retell” to the people what was decided at a higher level.
As for the high level, the dangerous part of this possible development is not even surveillance but the authority of “objectivity.” When an algorithm becomes the standard instrument, people naturally attribute rationality and neutrality to it, even when they know it is trained on selected data and works according to given criteria. And criteria are always political, only they are no longer conducted publicly.
Conclusion
The Genesis Mission can be read as an attempt to accelerate science and push the American industrial base forward. That is the “softer” interpretation, and it makes sense. But the second layer is more important because it explains why Genesis is also a political document.
In such a world, elections still exist, but they increasingly decide who will be the face of the system. It is much harder to democratically influence how the system calculates, what it considers risk, which goal it optimizes, and whose data it takes as truth. And that is what Genesis puts at the center. The state of the future may still be called a state, but sovereignty will increasingly hide in code, contracts, and data flows.
About the Author
