A: Now

  Human - Human - Human - Human -
    |       |       |       |
    AI      AI      AI      AI


B: Envisioned future

  AI - Human - AI - Human -
         |            |
        AI           AI

from the book

This is a critique.

Against the illusion of frictionless progress: A response to technological optimism

The only one around is a virtual section chief, who supports good ideas without getting in the way. It uses its strong generative AI skills to put together proposal documents, persuade the executive in charge, and even present to the board of directors. Wouldn’t that make for an open, comfortable workplace? (p.112)

When it comes to developing LLMs, we already have three major tech giants—Microsoft, Google, and Meta—so it will only get cost down. (p.136)

If AI eliminates language barriers, wages around the world will equalize. (p.142)

AI can serve as an excellent education tool, much cheaper than teachers and personalized. (p.154)

Isn’t an affordable, mostly remote AI university a good idea too? (p.175)

These remarks, coming from a highly respected sociologist in Japan, are deeply troubling. At first, I thought the author might be laying out an extended satire—one that would eventually be overturned or critiqued. But no: his argument, unfortunately, remains blind to the lived realities of work, education, and global inequality—even if we set aside the question of whether the kind of AI he envisions could ever actually come to exist.

Humans as economic residue

The idea that people will simply “move” to inefficient industries as automation proceeds treats humans not as individuals with agency, identity, or community, but as surplus labor to be absorbed wherever the market demands. This is the most daunting future I want to avoid living. Work should be a site of meaning and creativity, not an economic placeholder.

In addition, I would argue that, even if a person is not outperforming a machine, it doesn’t change the value of this person a bit. What is the meaning in outperforming a machine? Should we really compete to produce more and more, while we already know that the environment is not unlimited, and we’ve already out-produced its capacity? It is no longer the right phase for humanity to optimize and maximize. Some people already got more than they can consume in their life time, it’s high time we should think about inequality and cherish what we’ve got.

The bureaucratic fantasy of virtual management

The image of a “virtual section chief” who uncritically supports good ideas and efficiently handles proposals may sound appealing in a corporate context—but it erases the inherently human dimensions of judgment, dialogue, and dissent. [Hannah Arendt](https://en.wikipedia.org/wiki/The_Human_Condition_(Arendt_book)’s political theory reminds us that action is meaningful because it is unpredictable and plural, by automating managerial functions, we risk losing any of that we deemed human.

I would also add a grim observation that: in a society where labor has already become so alienated—where work evokes bitterness more than meaning, it’s no surprise that many might embrace solutions that promise to erase that pain altogether.

In Japan, where managerial roles are often associated with chronic stress, isolation, and moral exhaustion, the idea of removing that layer entirely may feel like a relief. This logic of erasure resonates disturbingly with the psychology behind train suicides: a desperate longing to escape a structure that has already crushed the self.

Interestingly, Japanese companies used to operate more like families. While hierarchy was respected, managers were more constrained compared to their Western counterparts—they had to be considerate and responsive to the voices of their subordinates to maintain authority. An employee’s value grew not through formal rank or qualifications, but through time spent building relationships, as personal connection was the most valued form of authority. (Nakane Chie, a vertical society)

However, with the flawed adoption of Western management styles, a contradiction emerged. Employees are still valued based on tenure rather than merit, yet managers have abandoned the virtues of consideration and proximity. They speak from a distance, and are emotionally and socially detached from the rest.

I can’t help but ask, isn’t the AI exacerbating this distance, instead of correcting it?

Flattening global inequality with AI?

Claiming that AI will equalize wages worldwide simply by removing language barriers is pure naïveté. I wouldn’t have believed it—if not for the fact that multiple Japanese people have written, and spoken that the hardest part of studying or working abroad is, without exception, the language.

But I would rather argue that, there is no perfect language skill, it’s the speaker, how their narrative is told, and in what context it’s told, that’s the most important factor contributing to progress and change. More broadly speaking, structural inequality is a result of histories of colonialism, capital accumulation, and geopolitical exploitation. The theories put forth by thinkers like Frantz Fanon and Thomas Piketty can be a valuable source to start thinking about the underlying economic and political structure, in which the technology is utilized.

It’s quite likely that AI technology will exacerbate current global inequality. Informational inequality is already prevalent, with AI being trained predominantly with English and the Western context. Non-English speaking countries have seen an increase in jobs related to localizing AI for a fraction of the pay as compared to employees in their English-speaking home country (but still highly rewarded when compared to the local jobs). Japan is experiencing the same thing, Google, Microsoft, OpenAI, and Amazon are dominating both the public and enterprise market, while Chinese investors and business owners are also biding their time.

AI is an aggregate of the most dominant information on the Internet (or arbitrarily chosen database), and when it becomes a dominant tool globally, I cannot help but wonder: How much local information is lost in translation, and pushed aside? and that is it true that there’s less need to learn and speak another language on your own? The threat posed by AI goes far beyond economics alone.

The AI-generated education

The idea of a low-cost education run by AI reduces education to content delivery. But education is not merely transactional, it should be relational, dialogic, and transformative. Personalization is not the magic remedy against institutionalized, sometimes exclusive educational systems. Can any part of what Maria Montessori embodied be automated? Paulo Freire also called education a practice of freedom, that, “The oppressed must be their own example in the struggle for their redemption.” (
Pedagogy of the Oppressed, 1970)

Maybe tools, or theories of education should be disseminated, not the contents itself.

Technological optimism in Japanese context

This particular strain of Japanese techno-optimism envisions technology as a mediator of harmony and hierarchy. Within this worldview, conflict is suppressed, and technology is tasked with smoothing over the messiness of human relationships. But as Ivan Illich warned in Tools for Conviviality, technologies that eliminate friction often eliminate freedom as well.

Conclusion: The cost of a frictionless future

These visions are, at their core, fundamentally inhuman. They ignore the purpose of life—not just efficiency or ease, but the value of struggle, of care, of becoming through multiple relationships. Still, the author offers a surprisingly detailed speculative blueprint, one that allows readers to vividly imagine this possible future unfolding in everyday work life.

Taken as a touchstone: how does this vision make you feel? Would you embrace the author’s envisioned future B?

One thing is clear, the current reality is no longer a place we can dwell in. To move forward, we must imagine and create a new context of humans working with AI.