The rise of artificial intelligence has been remarkably rapid. Its adoption has even outpaced that of the internet. In 1995, around 39 million people used the internet. Five years later, that figure had climbed to roughly 361 million, and the internet did not surpass one billion users until around 2005. ChatGPT, by contrast, reached 100 million active users just two months after its launch. The internet took years to reach the same milestone.
The comparison is not entirely straightforward. The internet was a new infrastructure that had to be built from the ground up. ChatGPT, by contrast, entered a world already connected by the internet, smartphones and social media. Even so, its rapid adoption demonstrates that generative AI was never merely a technological curiosity for specialists. Within weeks, it had reached millions of people, who began using it at work, in education and in everyday life.
This rapid expansion, accompanied by near-euphoric claims about the capabilities of large language models, has prompted a wave of speculation about how artificial intelligence will reshape society. As has often been the case since the time of the ancient Greek philosophers, the loudest voices tend to embrace the extremes.
One camp dismisses AI's significance almost entirely. According to this view, it is little more than an investment bubble, a marketing exercise and another overhyped technology that will eventually hit its limits. The other predicts a dystopian upheaval worthy of a post-apocalyptic film. Artificial intelligence, they argue, will deprive people of their jobs, their freedom and perhaps even their place in the world. In short, humanity is headed toward something resembling the world of the Fallout games.
Both positions share the same flaw: they make it harder to think clearly about what is actually happening. The first sees no change because the world has not yet been turned upside down. The second looks only to an imagined catastrophe in the future. One underestimates the present; the other sacrifices it to speculation about what may come.
A more balanced approach is therefore needed. That does not mean splitting the difference between optimism and catastrophism. The middle ground is not a compromise between two mistaken positions, but a search for what is true. We should recognize the revolutionary potential of this technology while also acknowledging the enduring nature of humanity, to which every technological revolution must ultimately adapt.
The Stock Market Shock
Artificial intelligence first sent shockwaves through the stock market, where investors price in the future long before it arrives. Nvidia has become the clearest symbol of that trend. In just a few years, the graphics chipmaker became the world's most valuable company, creating fortunes for countless investors along the way. Companies such as Palantir, meanwhile, have gone from relative obscurity to becoming widely recognized by investors.
Financial markets adapted to AI with remarkable speed. More importantly, they tied its future to unprecedented flows of capital. Technology companies are investing hundreds of billions of dollars in chips, data centers and energy infrastructure, while banks and investment funds are financing these ambitions and profiting from their growth.
The result is a self-reinforcing cycle. Rising expectations push share prices higher. Higher valuations attract fresh capital. Those investments are then cited as further evidence that the AI revolution is accelerating. The more money flows into the sector, the stronger the incentive becomes to sustain the narrative.
None of this means the AI revolution is illusory. The technology is real. But alongside it, a mythology has emerged about its almost limitless potential. Financial markets are no longer waiting to see whether AI succeeds. In many respects, they have already assumed that it will.
The Job Market Is in for a Long-Term Shift
The first cracks are beginning to appear in the labor market. This, after all, was where many expected the AI revolution to have its most immediate impact. Translators, copywriters and other content creators were widely predicted to be among the first casualties. That has not happened.
This is not because AI is absent from these professions. Quite the opposite. But producing more content does not automatically produce better content. Moreover, many of the professions considered most vulnerable have already adapted to working in highly competitive environments.
Being an influencer, for example, is not simply a matter of producing content. If only it were that straightforward. Success also depends on personal branding, reputation, the relationship with an audience and the ability to retain its attention. The job is far more complex than it first appears.
Artificial intelligence has not yet wiped out entire professions. What it has done is make life more difficult for people at the start of their careers. Junior employees were once given routine tasks that allowed them to develop their skills. Increasingly, those tasks can now be performed by machines.
Nor has AI delivered the dramatic productivity gains many expected. Ironically, the greatest disruption may ultimately affect the very people who built it: software developers. Programmers have long been among the most expensive employees for technology companies, making them an obvious target for automation. Even replacing a handful of developers can generate significant savings for a large business.
That represents a genuine shock, although only to those unfamiliar with history. Revolutions have always had a habit of devouring their own children. The labor market is undergoing a profound transformation, but it will be gradual rather than sudden. Its effects are likely to unfold over many years.
The same pattern can be seen across countless industries. Education is already experiencing an AI-driven transformation that has received relatively little attention, while healthcare may be next. Even here, however, the changes should not come as a surprise.
A True Social Shock
A genuine social shock must, by definition, be unexpected. That shock may emerge from the combination of content recommendation algorithms and generative AI.
The first part of that equation long predates ChatGPT. Social media algorithms do not present an objective picture of the world. Instead, they assemble a version of reality designed to maximize user engagement. Anger, fear and conflict therefore tend to dominate because they keep people on their screens for longer. The algorithms reward content that provokes outrage, anxiety or a sense of moral superiority, while measured explanations and nuanced counterarguments struggle to compete. Until now, they did not even need to create content themselves. They simply selected from material produced by others.
Generative AI changes that equation. Content is no longer constrained by human capacity to produce it. It can now be generated almost endlessly at negligible cost. This represents a far more significant development than a machine's ability to write an article or create an image. In the past, a propagandist had to craft a single message for a million people. Now, a system can generate a million different messages for a million different individuals.
This is where the real social shock begins. Artificial intelligence does not need to strip people of their free will. It only needs to keep showing them exactly what they want to see while gradually pushing everything else out of view.
Human beings have never experienced the world directly. They have always interpreted it through information filtered by others. But as more of that information is created by one machine and distributed by another, an entirely new environment emerges. Human nature remains the same. What changes is the environment's ability to exploit its weaknesses with unprecedented precision.
People Have Not Changed
People remain fundamentally the same. They still long for recognition, security, justice and truth. At the same time, they remain vulnerable to anger, pride, fear and complacency. New technologies did not create these impulses, but they can now monitor, measure and exploit them to an extent that was previously impossible.
The first major social shock caused by artificial intelligence may not have begun with ChatGPT at all. It began when algorithms assumed control over the distribution of information. Generative AI is merely the second stage of that process. It is the combination of the two that has the potential to reshape the very environment in which people think, make decisions and search for the truth.
The greatest danger posed by artificial intelligence, therefore, may not be that it will begin thinking on humanity's behalf. It began the moment it started deciding what people would think about.