Although eerie, there is also something almost comic in the way the heads of major AI companies talk about their own products. In a normal world, a company director tries to convince customers that his product is useful, safe, practical, perhaps even necessary. If he is selling a car, he talks about comfort and fuel consumption. If he is selling medicine, he talks about effectiveness and side effects, but carefully. If he is selling software, he talks about productivity.
And then the AI bosses enter the stage, still human beings, not androids, and tell us something roughly like this: "Our product could take your job, change the economy, destabilize society and perhaps lead us into a completely new phase of human history." Well then. Are you selling a subscription for the "pro" version or a ticket to the end of the world?
This is the strange paradox of today’s AI industry. The people who want us to trust their tools are, at the same time, trying to convince us that those tools are almost "too powerful" and dangerous for ordinary human life. On day one, they talk about "assistants" that will help us write emails faster, fine. On day two, they talk about systems that will automate a huge part of knowledge, work and perhaps decision-making itself. That pushes us straight into existential panic - but what if that is not accidental?
It is not hard to understand why all of this makes people uneasy. Who wants to buy a product from someone who sounds as if he is apologizing in advance for the damage that product will cause?
But the more interesting question is: why would anyone talk like that? The simplest answer is: because of money. Let us explain.
Destiny Has a Better Valuation
If you present artificial intelligence as a useful tool for programmers, lawyers and accountants, then you are building a large software company. Perhaps a very large one. But if you present it as a technology that will replace entire professions, reshape the labor market and become the infrastructure of a new world, then you are no longer selling a tool. You are selling destiny.
And here we come to the key detail. Destiny has a valuation. In other words, all this drama is designed to massively lure in investors. Look at it this way. If AI programs are truly being created that will, as their representatives tell us, change "everything", then where else should one invest but in "the end of the world"?
If AI is that important, then it must be funded immediately, massively and without too many petty questions.An investor who hears that a company is building a "smarter chatbot" might be interested. Maybe. But that phrase has already worn out. The market is flooded with chatbots. But an investor who hears that the same company may be building a machine capable of performing "most cognitive labor" in the economy hears something entirely different. He hears monopoly. He hears a platform more important than a search engine, more important than an operating system. He hears the possibility that a large part of the value of human labor could be transferred into ownership of models, data centers and licenses.
This is where fear becomes a financial instrument.
The greater the threat, the greater the promise. If AI can destroy jobs, then, clearly, it can also take over the market. If it can disrupt the economy, then it can also create a new economy. If it is dangerous, then it is important. If it is important, then it must be funded immediately, massively and without too many petty questions.
But that is not the whole story. It would be too simple to say that AI bosses are merely coldly performing panic in order to extract more money from investors. Some probably play that game consciously. Some play it half-consciously. And some, it seems, truly believe in their own drama.
California’s Playful and Apocalyptic Metaphysics
And this is where things become interesting.
Silicon Valley is not just a place where technology companies are launched. It is also a particular mental climate. For a long time, it has blended engineering faith in solutions, libertarian instinct, messianic capitalism and a very specific kind of fear of the future. From that world come "rationalism", "effective altruism" and the so-called x-risk circles, communities that have spent years discussing existential risks to humanity.
A Short Genealogy of Silicon Valley: Rationalists, Altruists and Prophets of Catastrophe
To understand why part of the AI elite speaks so dramatically, one needs to know three concepts from its intellectual underground: rationalism, effective altruism and x-risk circles. A note: here "rationalism" does not mean the classic philosophical rationalism of the 17th century, but a contemporary internet movement that emerged around forums, blogs, conferences and technology circles. It is not a mass movement, but rather a network of people who have spent years debating how to think "more rationally" than the rest of society.Rationalism starts from the assumption that people often think badly because they are full of cognitive biases. They believe what feels pleasant, underestimate rare dangers, overestimate their own intelligence and defend old views even when they no longer make sense. These technological rationalists therefore want to develop a colder, more precise way of making decisions, almost like mental software for correcting human errors. That can be useful. But there is also an unpleasant tone in this culture: the belief that social, moral and political problems can be solved if sufficiently intelligent people calculate them well enough.
From that world emerged effective altruism. The idea is simple: if you want to do good, do not be guided only by emotion. Calculate where your money, time or career can produce the greatest benefit. Better to save many lives with a cheap vaccine than fund a beautiful project that helps only a few people. On paper, reasonable. But over time, part of that movement turned toward very distant questions: not only how to help people today, but how to protect the future billions of possible people. This is where the big numbers, grand ambitions and slippery terrain begin.
The third concept is x-risk circles, from "existential risk". These are people concerned with threats that could destroy humanity or permanently close off its future: an asteroid, nuclear war, a pandemic, uncontrolled biotechnology, superintelligent AI. Their key logic is this: if a catastrophe is very unlikely, but its consequences would be enormous, we must not ignore it. This is precisely where much of today’s AI apocalypse was born.
The problem arises when these three worlds merge with Silicon Valley capital. The rationalist faith in one’s own clarity, the altruistic ambition to save humanity and the x-risk obsession with the end of the world ended up in the hands of people who also run companies worth tens or hundreds of billions of dollars. Then the story of AI is no longer merely technical. It becomes almost religious. Some build models, others warn of the end of civilization, others invest money, and together they speak as if the fate of humanity is being decided in a few conference rooms between San Francisco, London and Seattle.
This does not mean all these people are unserious, or that all their concerns are fabricated. Some are not. But their language often reveals something important. They do not see AI only as technology. They see it as a test of humanity, a moral mission, a threat and a promise in the same package. And that is why they sometimes sound less like directors of software firms and more like gatekeepers standing at the entrance to the future. That is where the problem begins.
Then ChatGPT appeared.
For most people, it was an impressive tool. A little strange, a little exciting, sometimes useful, sometimes foolish. But for part of the tech elite, it looked like a sign that their old obsession was beginning to come true. For years they had spoken of an approaching intelligence that might slip out of human control. And now, suddenly, a machine was speaking fluently, solving tasks, imitating personality.
Of course some of them got carried away.
Not because they were all frauds. But because the story was too seductive. If you spend your whole life imagining the moment when a dangerous new intelligence appears, then the moment the first convincing language model appears can feel like the beginning of a film you have already been watching in your head.
And suddenly you are not just the director of a company or the founder of a startup. You are someone standing on the threshold of the greatest change in the history of the species. Perhaps even one of the few who understands what is coming.
That is a dangerous amount of self-importance. It is difficult to carry without consequences.
That is why the apocalyptic rhetoric of the AI industry is not only marketing. It is also the language of status. A way of turning an ordinary product into a cosmic event. A way of turning a CEO into a prophet, an engineer into a savior, and a company into an institution that is not selling software but managing humanity’s transition into a new epoch.
Is that not interesting? The more dangerous the technology, the more important its creators sound.
Panic as a Fence Around the Market
There is a third layer as well, perhaps the most insidious one. Fear helps create regulatory advantage.
If an AI company says, "Our technology is very powerful and potentially dangerous", it is not only saying that risk exists. It is also indirectly saying this: that this technology should not be handled by small players, amateurs, open models, especially free ones, or unknown competitors. It should be led by "serious, large, well-funded companies like ours".
In other words, panic can serve as a fence around the market.
Clearly, large firms can always handle regulation more easily. They have lawyers, lobbyists, safety departments, links to the state and the money to comply with rules. Small competitors do not. So when an industry begins calling for "responsible regulation", one should always look carefully at who will ultimately profit most from that regulation.
Of course, this does not mean regulation is unnecessary. AI really does raise serious questions. Copyright, surveillance, war, propaganda, jobs, education, concentration of power. All of these are real problems.
But there is a difference between a serious discussion of risks and a theater of self-importance.
AI bosses often do exactly the latter. They speak as if they are burdened by the moral weight of the technology they nevertheless continue to push ever faster, while raising more capital, building larger models and seeking deeper entry into everyday life. They warn a little about danger, and immediately afterward ask for more chips, more energy, more data, more users...
That is the uncomfortable part of the story.
The public cannot endlessly listen to claims that this technology will destroy jobs, disrupt society and make people redundant.Because if someone truly believes they are building technology that could seriously damage society, one would expect them to slow down. To stop. To say: perhaps we must not leave this to the market until we know what we are doing.
But that, of course, is not happening. Instead, we get a strange combination of confession and sales pitch. "This could be terrible, but we are the best people to build it."
That sentence, in various forms, is the heart of today’s AI politics.
But it seems that this strategy is beginning to crack. The public cannot endlessly listen to claims that this technology will destroy jobs, disrupt society and make people redundant while still retaining warm trust toward the people selling it. At some point, fear stops being fascinating. First it becomes fatigue. Then resistance. Have we already reached this last phase so quickly?
And can you guess where the rhetoric of these people is going now?
Exactly. We are now watching a change in tone. Suddenly there is more and more talk of "augmenting" people, not replacing them. Of tools that expand capabilities, not machines that take over the world. Of the fear of job loss being "perhaps exaggerated".
That does not mean they changed their minds out of pure nobility. More likely, they realized something more banal: people do not like companies that openly threaten them.
Especially not when those companies want to enter schools, offices, hospitals, government systems, personal devices and daily habits. It is difficult, what a surprise, to be at once an infrastructure of trust and a source of apocalyptic announcements.
And then there is Wall Street. While you are a startup with a big story and the mystique of the future, it is useful to speak like a prophet. Perhaps even an apocalyptic prophet. But once you approach public markets, major institutional investors and more serious oversight, that prophetic tone begins to sound messy.
Less Terminator, More Excel
That is why it is possible that we are now entering a new phase of AI rhetoric. Less Terminator, more Excel. Less "everything will change", more "we will help employees become more efficient". This is calmer language, more acceptable to the public, regulators and large buyers.
But we should not fool ourselves. Beneath the softer rhetoric, the same question remains: who will own the new infrastructure of work?
That is the most important trick. Not to convince us that machines already think like humans. But to convince us that the future has already been allocated.Because the real issue is not only whether AI will take jobs. That matters, of course. But the even more important question is this: if AI increases productivity, who receives the surplus? Workers, users, society or the owners of the models?
In other words, even if the apocalypse does not arrive, the concentration of power can arrive very quietly. Without spectacle. Without robots marching through the streets. Simply through licenses, data centers, contracts with governments and the quiet transformation of knowledge into a service that a handful of companies rent out to the rest of the world.
That may be less cinematic. But politically, it is more important.
That is why one should be careful with the grand stories of the AI industry, both when they frighten and when they reassure. Both versions can serve the same goal. In both cases, they want to be inevitable.
That is the most important trick. Not to convince us that machines already think like humans. Not even to convince us that AI will soon replace all labor. But to convince us that the future has already been allocated. That it is normal for a few companies to speak on behalf of an entire civilization. That it is natural for their CEOs to discuss the fate of work, society and human purpose as if they had received a mandate from history.
They have not.
AI should be taken seriously. It is not a toy. Nor is it merely a passing hype. The technology is already powerful enough to change working habits, education, entire industries and the way we make decisions. Rules are needed. Public debate is needed. Skepticism is needed even more.
But we do not have to accept the drama on the terms offered to us by the people who profit most from that drama.
Perhaps AI will change the world. It probably will, at least in some form. But the question is not only what the technology can do. The question is who owns it, who controls it and who is allowed to say: this is what the future will look like.
Everything else is smoke, marketing and that Californian metaphysics.
And realistically, we have already heard more than enough of it.
Comments