Guardian columnist Zoe Williams: conversations about AI are no longer keeping pace with the technology itself

Guardian columnist Zoe Williams describes a strange state in which public discourse about AI finds itself: everyone talks about it, but almost no one keeps up with how quickly the technology itself is changing. As a result, even the loudest debates sound as if they had become outdated before they even ended.

Topic Fatigue

The text begins with an everyday scene: at a celebration of the author's 80th birthday, the author meets a scientist who doesn't immediately admit that he works in computer science. The reason is simple and almost comical: he doesn't want to start another conversation about artificial intelligence. For Williams, this instantly sets the tone for the entire column.

AI is debated so much that the very topic has begun to tire even the people closest to it. According to her thinking, the problem is not only in the quantity of conversations, but also in their quality. Almost any position—enthusiastic or apocalyptic—fails to keep pace with reality.

While someone repeats that new technologies always frighten at first and then cure cancer and advance civilization, the technology itself has already managed to change its scale, interface, and field of application. Because of this, public discussion increasingly sounds like a play about the future written from yesterday.

Speed vs Reflection

The main point of the column is that AI develops faster than society manages to comprehend it. One can demand freedom for developers, one can insist on strict regulation, but both positions often lag behind. While people discuss what exactly should not be entrusted to machines, the models are already becoming better at performing that very task. Therefore, ordinary exchange of opinions begins to feel not like control, but like capitulation before the pace of change.

"We will be destroyed by a craving for endless chatter that leads nowhere."

Williams does not argue that discussion is unnecessary at all. On the contrary, she acknowledges: reading, understanding, and debating are necessary. But here too there is an unpleasant paradox. We are accustomed to considering conversations as a way to collectively develop a solution, but in the case of AI they constantly stall, break down into camps, and do not give the sense that someone is actually holding the situation in hand. Even thoughtful discussion loses to the speed of a system that reads, summarizes, and changes faster than a human.

Where Risks Are Visible

The author lists not one universal threat, but a whole set of lines of tension. Criticism of AI, according to her version, covers both philosophical fears about the end of human creativity and very applied questions of employment, power, and infrastructure. Therefore, the debate comes down not to a choice between "technology will save the world" and "technology will destroy everyone," but to many specific consequences that are already beginning to manifest.

• pressure on service professions and routine office work
• reduction in entry-level positions for young professionals
• automatic filtering of resumes and applications before contact with a human
• environmental cost of computing and data centers
• military and politically repressive applications of AI

The most unpleasant thing in this picture is that all these questions require urgent attention, but the mechanisms of public reaction themselves remain slow. One can discuss as much as one likes the interests of billionaires controlling key AI companies, or argue about the real scale of environmental damage, but the feeling of lagging behind never disappears. By the time the debate becomes substantive, the subject of the debate has already managed to change.

What This Means

The Guardian column does not offer a program of action, but certainly captures the mood of 2026: society is tired of talking about AI and at the same time cannot stop. For the industry, regulators, and ordinary users, this is a signal that the previous pace of discussion is no longer sufficient—a faster, more applied, and more substantive conversation is needed about where technology is genuinely useful and where it needs to be limited before it moves forward again.