Habr AI on the future of work: how AI and robots could return society to a new antiquity

Habr AI published a reflective column about how the AI era could transform society more deeply than past technological breakthroughs. While machines previously changed mainly the tools of labor, neural networks and robots are now beginning to take over the executor function itself.

Four Forms of Intelligence

The author proposes viewing the future as an environment where multiple types of intelligence coexist simultaneously. The first is ordinary people. The second is AI systems based on LLMs or their future successors.

The third is cyborgs—people with neural implants and other means of direct cognitive enhancement. The fourth option still seems almost fantastical: digital copies of consciousness that can continue to exist separately from a biological body. From this scheme, at least the "human plus AI" combination already works in reality.

It is precisely this, the author argues, that has begun to rapidly reshape the market for knowledge and services. Professions where a significant part of the output can be represented as text, analysis, translation, or standard solution face pressure first. Translators, editors, support staff, some analysts and office workers increasingly compete not with another human but with a model that only needs the task to be properly formulated.

Why This Is Different

In past technological revolutions, professions also disappeared, but the logic was different. Coachmen were replaced by train drivers, then drivers; trades were automated, but the final labor remained human. A new machine changed the qualification, not the principle of employment itself.

A person learned to work with a steam engine, a machine tool, or an automobile, but remained the center of the production process and the recipient of the primary economic role. With AI, according to this logic, the shift is more radical. Now automation can take on not only physical effort but also a significant portion of cognitive work: searching, summarizing, generating drafts, preparing solutions, and even dialogue with clients.

The human role shifts toward formulating the request, checking the result, and controlling the system. This means that value could concentrate among those who control infrastructure, data, and access to automation, rather than among the broad market of executors.

Robots and a New Class

The author believes this shift won't stop here. If sufficiently skilled robots are added to neural networks, the impact will extend beyond just office work to physical labor as well. When a machine can cook food, replace a faucet, clean an apartment, deliver goods, and eventually service machines like itself, the amount of conventional work will shrink dramatically. Then the question will shift from productivity to distribution: who earns income in a world where the main tasks are performed not by humans. From this scenario, the column draws several harsh conclusions:

• the mass labor market could shrink faster than people can retrain;
• managing AI and robots will require a small layer of highly qualified specialists;
• states will face pressure in favor of basic income or new redistribution schemes;
• military, security, and critical infrastructure will also become increasingly automated;
• political and economic power risks becoming even more concentrated at the top.

"The top of society will never give up power."

Hence the association with antiquity in the headline. It's not about a literal return to the past, but about the possibility of a new class order where a narrow elite controls the means of production, access codes, and automatic systems, while most people lose their former economic indispensability. Even if the scenario turns out to be less harsh, the very question matters: the technological race increasingly looks less like simply updating software and more like a dispute over the future of social structure.

What This Means

The Habr AI text does not provide an exact timeline forecast, but it fixes the main nerve of the debate around AI: the problem is no longer reducible to model quality. The key question is how societies will divide income, power, and responsibility if both mental and physical labor begin to massively transfer to automation.