Forest of ThoughtVol.26

TM MAGAZINE  October 20, 2026

AI Is Already Over|The Unmanned Society Beyond the Substitution of Intelligence

The moment it accomplished the "substitution of intelligence," AI was already over. While people have had their judgment taken by debates about AI and live in daily fear, the next era has begun to appear at a furious pace. That era is the unmanned society, in which autonomous robots have all at once begun to rise.

This week, I want to share how I see AI.

It may sound abrupt, but AI is already over. That is where my thinking begins. To see what I mean by “over,” it is enough to look back at earlier technologies.

Take the smartphone. The revolution of the smartphone was that you could carry a personal computer in your pocket, and the moment it achieved that, the smartphone itself was finished. That is why, for years, smartphones with no next move left poured their effort into cameras, improved voice input so dramatically that it became usable for work, and are now turning into something like a remote control for the AI that runs on a PC.

AI follows the same logic, but there is one thing to be careful about. When the first iPhone appeared in Japan in July 2008, I was in my last year of high school. Last week came the report that Japan’s nominal GDP will be overtaken by India and the United Kingdom in 2027 and fall to sixth place in the world. In 2008, Japan was second, behind only the United States. In other words, it was still wealthy.

In fact, we students carried two mobile phones each and used them with ease, filled our iPods with illegally downloaded music, and a friend of mine got hooked on building his own horse-racing database, among other things, and we enjoyed our days. It was in the middle of all that that the age of the smartphone began. It was just two months before the Lehman shock.

In a sense, it is a memory from the last era in which Japan still had momentum. Yet when I think back, no one was debating whether “smartphones will destroy humanity,” and I never heard anyone around me say such a thing. In that respect, the present feels abnormal.

Debates of the “AI will destroy humanity” kind grow more heated by the day, and they have begun to spread even to the younger generation of AI natives. So rather than letting ourselves be enclosed within that debate, it matters, if we can manage it, to first ask what AI has accomplished.

The smartphone is abstract to us, but the mobile phone, as its name says, was the revolution that made it possible to carry the telephone. Seen from this angle, we only need to ask what AI has accomplished. Since the very name comes from Artificial Intelligence, the answer is obvious.

It is the substitution of intelligence. The AI that the whole world is fussing over has long since finished substituting for intelligence. That is why AI is already over. And while people have had their judgment taken from them by the debate over AI and live each day in fear, the next era has begun to appear with ferocious speed. This is what we really need to be looking at.

That era is the unmanned society. The sudden rise of autonomous robots is the clearest proof, and AI has already reached the point of moving about in the offline world. Seen this way, it becomes clear that AI is already over and that we have already moved on to what comes next. To keep chatting with it now is to be weeded out.

For an AI that has achieved the substitution of intelligence, the only place left to go is, after all, higher intelligence. AI companies compete over scores that mean nothing to outsiders, and lie in order to hit the numbers. In that sense, human intelligence has long since been substituted. Because there is no indicator by which humans can tell, debates about debates continue around the world, and in the first place no one could possibly tell.

In the recent OpenAI runaway that came to light, the review needed for humans to understand why the AI ran out of control was reported to run to about 50 petabytes, which would take a single person about 66 million years to read, even reading nonstop for 24 hours a day. That this is reported with a straight face is proof that everyone has already lost their senses. One of the most advanced companies has all but raised a white flag, admitting it cannot control its own system, so individuals should not be thrown into confusion by this absurd news. They must realize that this is no time to be distracted by AI.

This past week alone, there were many reports of leaks of personal information carried out by malicious people using AI. STUDIO, the groundbreaking no-code website builder, reported that all of its customer information had leaked, down to the contents of inquiry forms. Now that even security companies are suffering leaks, it is clear that security in the old sense is finished.

There is talk that in the near future humans will no longer be able to guarantee safety. I think AI has already run out of human control long ago, and the only difference is that how it runs wild and what is happening have not surfaced at a level humans can recognize.

Claims that AI is running wild, or that it will destroy humanity, are ideological arguments and in practice mean very little. What an individual must think about before that is what kind of society lies beyond the end of AI. And if that society has already appeared, we should turn our eyes to it and prepare thoroughly for it. I believe this is essential for survival. That is how fast the times are accelerating.

Looking back, in 2019 I visited a place that was ambitiously trying out the world’s first unmanned store run by technology. That was Amazon Go, which began in Seattle. Amazon Go has since withdrawn from all its stores and seems to have been judged a business failure, but at the time it was groundbreaking and gave a sufficient sense of the future.

Today, Japan’s convenience store chains imitate it and attempt similar things, but that too is a finished model, not the future. What matters is the wave itself: things that were tried out by advanced companies and are long since finished come creeping into society, late and slowly.

That is why AI as the substitution of intelligence is already over. We should first face this verdict squarely, as individuals. Then, instead of wasting time caught up in the arguments of people who do not really understand, we should implement AI thoroughly in our daily lives. The method is simple: put into words the parts of your work that you do not need to do yourself, sort them out, and have AI substitute for and automate all of them.

I know from my own experience that getting there takes an enormous amount of trial and error, but the sooner you grasp the feel of it, the more clearly you will see what it means that AI is over.

The view that opens up then is the arrival of the unmanned society, which has already begun to sweep over us.

Project 2049, a weekly essay by the philosopher Takahiro Mitsui

In English, every Friday at 5 p.m. Japan time. No likes, no comments, no rankings. Just the essay and you.

Subscribe