12 Key AI Trends That Will Rewrite the Rules of the World: Money, Power and the Future of Humans
Which changes in artificial intelligence are truly fundamental: who will get the main advantage, where money and power will go — and what should you personally do about it?
OpenAI has more than 1.2 billion weekly users, and that scale gives the company the right to make mistakes and lose money, so it is better, in Alexander Volchek's view, to use the biggest system. People will pay for AI not like utilities but like a car or a mortgage if it helps a thousand times more. Sam Altman admits that 'some bad things will happen', and no human can check what AI can check. Apple is hopelessly behind, Meta is shutting down Limitless, which it bought for hundreds of millions of dollars, and video and voice can no longer be taken at face value. AI will not replace the living human: no program will love a child in a parent's place. Only the US and China are left in the race, money in AI is growing toward trillions, and regulation cannot keep up. Whoever has chips, data centers and energy wins — but for a thousand dollars a month you can now build systems that three years ago would have required hiring people, and results are worth counting over the long run.
What to watch for
Key takeaways
OpenAI has more than 1.2 billion weekly users — the right to make mistakes, to infinite money and to not caring about profit. Contact with a billion people through AI goes deeper than any messenger or social network, so Alexander advises using the biggest system; Anthropic, he says, is building superintelligence, not a chat for you.
Twenty or two hundred dollars a month for AI seems expensive, although people easily buy cars for tens of thousands of dollars. If AI helps a thousand times more than a car, you can pay for it like a mortgage or a lease — and people's budgets will be rebuilt.
To create a system smarter than humans, it has to be let go — and it is being let go. Sam Altman said AI should not be restricted and that the world must accept that 'some bad things will happen' — although a few weeks earlier he had said the opposite.
Apple devices hold more than half of the US market, but the company has lost the AI market so badly that Alexander questions whether it will survive: this is how the mobile platforms of earlier players once disappeared. Companies like OpenAI will have to become an operating system.
Claims that a new model will replace someone can be laughed off: the leaders will buy any player in AI and destroy it. Meta bought Limitless, which made a recording device, for hundreds of millions of dollars and is now shutting the project down.
Alexander has voiced this thought eighteen times in eleven episodes. Finances are an incredibly personal subject, but the more you share, the more new scenarios you get, and OpenAI, Anthropic and Google have some of the best security in the world.
AI already surpasses every system for detecting fake voice and video, and recognizing fakes will become ever more impossible — it is only a question of money. Meanwhile we can no longer check almost anything serious: how do you check Altman's statement — over a year or over fifteen years?
You will never vacuum a room the way a 'puck' of the same power does: it runs on a schedule, sees stains and recognizes crumbs with sensors, and Alexander's vacuum works about five hours a day. People keep comparing the incomparable.
Alexander's eldest daughter is learning to drive, but his three younger children most likely won't need to: a CyberCab with no steering wheel is on the roads and will cost about two hundred dollars a month. A car with cameras is safer than a human and will free people from traffic jams.
If you treat a person as a material object, AI will replace them, but a person has a soul and a spirit. Thinking that AI will love your child is as foolish as believing the brain knows more than a computer; an avatar talking to the children Alexander calls madness.
Europe, Russia and everyone else dropped out more than two years ago — in resources, chips and infrastructure. China is held back by the lack of global platforms and distrust of its privacy, and what matters more than countries are the companies, the lobbyists and the agreements between them.
Silicon Valley entrepreneurs valued OpenAI at $50–100 billion, while Alexander spoke of ten trillion. The economics of AI, he believes, do add up: WhatsApp, YouTube and Telegram were long unprofitable too, and OpenAI and Anthropic will go public at valuations of several trillion.
For the last hundred and fifty years laws have failed to keep up with what is happening — COVID is an example — and AI's pluses and minuses are far bigger. Yet with AI one can now build even high-load systems, and Anthropic's Mythos finds problems on tens of thousands of sites.
These are not people but groups of political structures. But it is still possible to break in now: for a thousand dollars a month serious systems get built, while three years ago that money would only have hired an entry-level marketer, salesperson or programmer.
What should be counted is real results, money and time — not pretty demos and short-term profit. Teaching in 2026 how to change the screen resolution in Windows 10 does not develop the brain: it did in 1993–1994, but not now.
What this episode is about
A ToTheMoon episode that is the third part of a series about the main trends in artificial intelligence. After ten rules and ten trends, Alexander Volchek adds twelve more — on money, power and the future of humans — and numbers them from zero to eleven. The twenty thoughts of the previous episodes were collected by a system by how often they were mentioned across all of the channel's more than one hundred and sixty episodes, and all of them are published on the project's website; Alexander also often raises the new topics in different episodes and runs into them constantly himself.
The first trend is that OpenAI has the strongest models and a huge lead, although the company lags behind on certain features and interfaces. It has more than 1.2 billion weekly users, and that scale gives it the right to make mistakes, to infinite money and to not caring about profit, so it is better, Alexander advises, to use the biggest system. Anthropic is building superintelligence that will create businesses and software itself rather than a chat for people, and Google, the leader only a year ago, is awaiting a new version of Gemini and may yet burst back into the top layer.
Trend zero is that people will pay for artificial intelligence not like utilities but like a car, a robot or a mortgage: if AI helps a thousand times more than a car, twenty or two hundred dollars a month is not expensive. The second trend is that AI is becoming autonomous and faster than it can be controlled: a system smarter than humans will have to be let go. Sam Altman admits that 'some bad things will happen', and no human is able to check what artificial intelligence can check.
The third trend is that Apple is hopelessly behind, and Alexander questions whether the company will survive: this is how Windows and the mobile platforms of earlier players once lost their markets. Companies like OpenAI, he believes, will have to become an operating system, and xAI wants to rename itself ASI following the US government's new wording. The market is closed to most: Meta bought Limitless for hundreds of millions of dollars and is now shutting the project down.
The fourth trend is to hand artificial intelligence your finances, taxes and bookkeeping: the more you share, the more scenarios you get, and the biggest systems have some of the best security in the world. The fifth is that video and voice can no longer be taken at face value, and a fake made by AI can no longer be told apart. Alexander goes through the story of a viewer who asked AI how qualified he is, and asks in return: does AI know whom it is judging and whom it is answering?
He calls the dispute about the robot vacuum a comparison of the incomparable: the 'puck' drives around the house five hours a day and sees stains with its sensors. The sixth trend is that AI will not replace the living human: Alexander's eldest daughter is learning to drive, while the younger ones most likely won't need to, with a CyberCab with no steering wheel on the roads — but neither a program nor an avatar will love a child in a parent's place.
The seventh trend is that only the US and China are left in the AI race, while Europe and Russia dropped out more than two years ago; China is held back by the lack of global platforms and distrust of its privacy, and what matters more than countries are the companies, the lobbyists and the agreements between them. The eighth is that money in AI is growing toward trillions: Alexander has long said such companies will be worth ten trillion each, and he disagrees that their economics do not add up — WhatsApp, YouTube and Telegram also brought no profit for a long time.
The ninth trend is that regulation slows things down more than it helps, and governments cannot keep up. The tenth is that whoever has chips, data centers and energy wins. But it is still possible to break in now: for a thousand dollars a month, systems get built that three years ago required hiring people. The eleventh is to count real results, money and time over the long run, not pretty demos.
The third part of the series shifts from personal skills to who owns the future: money, compute and trust. The sharpest claims — that Apple may not survive, that AI is developing faster than it can be controlled and that only the US and China are left in the race — Alexander backs with examples from his own life: Limitless, bought and shut down, the robot vacuum, his daughter learning to drive while a CyberCab is on the roads. The episode's practical conclusion is to break in now, while it is still possible, and to count results over the long run.
Episode transcript
The episode is in Russian; below is an English reading guide to the transcript (the full EN transcript is a machine translation). Voice matching applied to 90 segments: 90 identified, 0 mixed, 0 probable, and 0 unresolved.
Read transcript on a separate page
Loading…