Many people first realize the terror of AI not because it can code, draw, or make PowerPoints.

It's that day when, faced with a slightly complex problem, your hand reaches for ChatGPT or Claude before your brain can even begin to unfold. An opinion hasn't yet formed, but AI has already laid out a structure for you. A report hasn't been fully read, but AI has already told you the key points, logic, and conclusions.

AI saves you the effort of thinking, but it also takes away that clumsy, slow, painful, yet necessary process before arriving at an answer. The hesitation, trial and error, detours, wandering in dead ends, and finally, that moment when two clues suddenly connect—all of it is collapsed.

This feeling is now being increasingly validated.

Scholars Steven Shaw and Gideon Nave from the Wharton School of the University of Pennsylvania have precisely named this subtle psychological state "Cognitive Surrender."

In their 2026 study, "Thinking, Fast and Slow and Artificial Intelligence: How AI Is Reshaping Human Reasoning and the Rise of Cognitive Surrender," the two researchers propose a "Tri-System Theory." In the traditional dual-system model of cognitive psychology, System 1 represents intuitive, fast responses, while System 2 represents deliberate, slow logical reasoning.

Now, AI is forcefully intervening, becoming a third, external artificial cognitive system operating outside the brain: System 3.

They conducted an experiment. 1,372 participants were asked to complete 9,593 reasoning problems similar to the Cognitive Reflection Test (CRT). The peculiarity of these problems is that they often have a highly tempting intuitive answer (designed to make System 1 err), requiring participants to actively engage in deeper thinking (System 2) to override their intuition.

In the experiment, using AI assistance was optional, and the results showed that participants actively chose to consult AI on more than half of the problems.

The researchers randomly controlled the accuracy of the AI's answers through hidden prompts in the background. The result was that once AI assistance was enabled, 90% of participants followed the AI's correct suggestion. Yet, when the AI provided a wrong answer, a staggering 80% of people still chose to follow it blindly.

Here, this double suppression of both intuition and deliberative thinking is precisely the behavioral hallmark of 'Cognitive Surrender.'

AI not only answers for you, but it also makes you relinquish your own judgment on 'whether you need to keep thinking.'

01

Human Civilization Is a History of Outsourcing Abilities

If we zoom out the historical lens, we will find that AI is not the first tool to plunge humanity into the panic of 'losing abilities.'

In fact, outsourcing abilities to external mediums is the very foundational logic of human civilization's evolution.

In ancient Greece, Socrates vehemently opposed the spread of writing, arguing that recording knowledge on parchment scrolls would destroy the internal muscle of human memory.

This wasn't entirely unfounded paranoia.

In 2020, Louisa Dahmani and Véronique Bohbot published an in-depth study on the correlation between GPS use and spatial memory in Scientific Reports. They tracked the GPS use and spatial memory of regular drivers. The results showed that the more habitually one uses GPS, the worse one's spatial memory is when navigating without it. The researchers also cautioned that this isn't simply a case of 'people with a poor sense of direction using GPS more,' but more likely that GPS dependence itself causes spatial memory to atrophy.

GPS didn't destroy humanity's overall intelligence, and the popularization of navigation apps has even dramatically improved the efficiency of logistics and travel for the entire society. But it did irreversibly weaken the specific human ability to build cognitive maps of the physical world.

But atrophy doesn't mean elimination; it signifies a transfer.

In 2011, Betsy Sparrow and her collaborators at Columbia University published a paper in Science titled "Google Effects on Memory: Cognitive Consequences of Having Information at Our Fingertips."

Through four meticulously designed experiments, the research team discovered a phenomenon. When people are faced with difficult questions, their brains are primed to subconsciously think of computers and search engines. Moreover, when people know subconsciously that a piece of information can be easily found online later, their probability of remembering the content of that information drops significantly; instead, their memory is strengthened for 'where and how to find that information.'

The search engine did not erase memory. It changed the form of memory.

Humans shifted from remembering content to remembering entry points.

From antiquity to the present day, it is precisely this instrumental form of outsourcing and ability disarmament that created a driving force for civilization's advancement.

Humans outsourced digestion to fire, cutting, carrying, and striking to tools, memory to writing, calculation to calculators, and trust and cooperation to rules and institutions.

Humans didn't become civilized by keeping all their abilities stored within their bodies and brains. On the contrary, it was by continuously migrating abilities to the outside world—to tools or to the group—that more complex civilizations were constructed.

So isn't this a good thing? We can just let tools continue to improve.

But years of research in sociology and philosophy prove this process is not without cost.

Its greatest cost is human alienation and the deepening malaise of our post-industrial age.

02

The Historical Cost of Cognitive Surrender

From the process of outsourcing abilities just described, we can see that as the ability to remember specific content was outsourced to tools, people began to train themselves to search efficiently.

Whatever a tool takes over, the old ability fades away. People also reshape their abilities into new forms.

If tools save us from certain abilities, where does this freed-up cognitive bandwidth end up?

It shifts towards depth and fragmentation.

Skill Outsourcing and Human Alienation

In 1974, Harry Braverman, in his classic work Labor and Monopoly Capital, dissected the phenomenon of 'deskilling' within the capitalist labor process.

Braverman pointed out that one of the core actions of modern management science (like Taylorism) is the cruel and complete separation of the 'conception' of labor from its 'execution.'

In the pre-industrial era, a blacksmith or carpenter held mastery over the complete lifecycle of an object, from design and material selection to forging and polishing. Knowledge and action were unified within the craftsman.

But with industrialization and the advance of monopoly capital, this complete process was broken down into countless fragmented steps that could be precisely measured, standardized, managed, supervised, and replaced at any time.

Our cognition, as a result, has been splintered into narrow directions. The rest is either outsourced to tools or to other people.

Wholeness no longer exists. This is, in essence, alienation in the Marxist sense.

For Marx, labor was not just a means of livelihood. Through labor, a person objectifies their intentions, judgment, skill, and imagination into the world. When a person makes a table, a chair, or an object, they don't just acquire a product; they also confirm their own capability within it: I understand the material, I can organize the process, I can turn an idea into reality.

But when labor is split apart by an extreme division of labor, this circuit of self-confirmation is broken. The laborer no longer faces a complete creation, only a small, measured, supervised, and replaceable set of actions within the process. They participate in production but no longer truly own the production process; they create value but increasingly struggle to know what this value has to do with themselves.

Thus, the product is no longer an externalization of a person's ability but becomes an alien thing. It belongs to capital, to the factory, to the market, to the management system, but no longer seems like the laborer's own life activity. A separation emerges between a person and their creation, mediated by wages, processes, machines, management, and ownership.

Braverman's critique of Taylorism pushes this alienation deep into the labor process itself. Capital seeks not only to own the product but also the knowledge of 'how to produce.'

Conception is taken by management, execution is left to the worker. The laborer thereby loses not just the complete product but the overall command of their own abilities.

The necessary steps to complete this outsourcing are the very acts of ability outsourcing and cognitive surrender. When the results produced come to represent me, evaluate me, and dominate me, yet I can no longer confirm myself within them, it becomes alienation.

And this alienation has been continuously deepening along the trajectory of the 20th and 21st centuries.

The division of labor begins to extreme toward one side

By the latter half of the 20th century, with the rise of computer technology, the fragmented niches society required people to fill underwent an even more polarized shift.

In 2003, David Autor, Frank Levy, and Richard Murnane from MIT published a highly influential paper, "The Skill Content of Recent Technological Change: An Empirical Exploration," in The Quarterly Journal of Economics.

Through analysis of massive labor market data from 1960 to 1998, this study posited a core judgment: the substitution of human labor by computer technology (as a form of capital input) is not indiscriminate. It ruthlessly replaces 'routine cognitive and manual tasks,' i.e., procedural work that can be broken down into explicit rules and expressed in algorithms and code.

However, at the same time, computer technology strongly complements and demands 'non-routine analytic and non-routine interactive tasks,' such as complex communication, solving ambiguous problems, and interpreting technological outputs.

Autor's team's findings imply that technological outsourcing in the mid-to-late 20th century didn't stop people from working; instead, it pushed people into tasks that are more abstract, more unstable, and harder to leave.

You no longer manually organize data, but you must interpret it. You no longer memorize a route, but you must judge the platform's rules.

The rise of tools has made more abstract, more institutionalized, and harder-to-exit thinking the core of our skill development.

This has led to two familiar ailments of our era: one is exhaustion, the other is a loss of security.

First, exhaustion. More complex work inevitably brings greater depletion, and with this shift, the low-load segments of the workday are continuously siphoned away. Those boring, repetitive, mechanical, seemingly inefficient tasks once provided a natural cognitive buffer between complex labors. They allowed people to temporarily pull back from high-stakes judgment and complete still-necessary work at a lower power consumption.

When human labor is increasingly narrowed down to a continuous stream of complex cognitive tasks, this buffer disappears. You no longer alternate between light and heavy tasks but jump from a meeting requiring judgment, to data that needs interpretation, to an email requiring a careful tone, and finally to a decision with consequences to bear.

This is precisely what Hartmut Rosa describes as 'social acceleration' manifesting in daily work life. Technology promises to save time, but the time saved does not naturally become leisure; it is instead refilled with higher-frequency tasks, faster responses, and denser communication.

Exhaustion is generated right here. Not because every task is overwhelmingly difficult, but because nearly every task demands you to stay alert, understand the context, make a judgment, coordinate with others, and be accountable for the outcome.

Secondly, there is a loss of security. As the division of labor continuously refines itself, the definition of an individual's value is also sharply narrowed. Whether you are worth anything increasingly depends on your ability to continuously output measurable results within a specific local system: Can you take more orders? Can you respond faster? Can you write a better-looking weekly report? Can you adapt to more tools? Can you perform as an upward growth curve within an evaluation system?

A person's capabilities become tethered to increasingly narrow scenarios. One is either forced to remain completely flexible, ready to accept low-skill, low-security, replaceable tasks at any time, or pinned down in a highly specialized position, industry, or set of platform rules, betting all of one's training on an ever-shrinking value coordinate.

Zygmunt Bauman, in Liquid Modernity, describes precisely this collapse of security. In earlier industrial society, a person might have been firmly fixed by a factory, a work unit, a class, and a profession, but this fixity at least provided predictability. You knew what bound you, and you knew what you relied on to survive.

Now, you can only live and die with an industry, or even a single job role itself.

Unfortunately, the speed at which jobs are being swallowed by technology is accelerating. Mobility becomes almost inevitable.

21st Century: The Narrowing to the Interface

As time moved into the 21st century, this narrowed, domesticated ability to adapt to social evaluation systems has even narrowed down to the interface layer.

In 2016, tech ethnographers Alex Rosenblat and Luke Stark published a study on Uber drivers. In their research, the emergence of ride-hailing platforms like Uber fundamentally restructured a driver's skill tree. As Rosenblat and Stark found, their new skill became reading the platform. When to go online, where to wait for rides, whether surge zones are reliable, which orders are unprofitable, how ratings affect future dispatch—these elements constituted the new street map.

These forcibly narrowed new skills became a type of extremely parochial 'platform-dependent skill,' highly attached to a specific digital environment.

This is more hopelessly fragile than a niche industry itself.

A carpenter from the pre-industrial era possessed an understanding of wood grain and a physical intuition for mortise and tenon structures. This skill was personal and oriented towards the entire material world. Even if he changed villages or toolsets, as long as wood existed, his skill would forever be valid and could always be exchanged for survival resources.

But a food delivery rider proficient at 'gaming the system,' or a ride-hailing driver extremely good at reading dynamic pricing—the complex skills they pride themselves on are entirely dependent on the servers of a few tech companies.

This skill possesses no trans-platform transferability.

Throughout the 20th and 21st centuries, alienation, facilitated by this technology-driven outsourcing, has sprinted towards an even narrower, more fragmented self.

Taylorism took the production process away from the person, computerization took routine tasks away, and platformization took the operating environment away.

But in each round, people were still forced to form some new ability in the remaining position.

What AI brings, however, is not a continuation of this direction, but a brand-new form of alienation.

03

The New Alienation of AI

If Taylorism compressed people into hands within a production flow, computerization pushed people towards more abstract interpretation and coordination, and platformization trained people to become adapters within algorithmic interfaces—after each round of outsourcing, people were still forced to form some new ability in the remaining position.

Writing weakened memory but strengthened reading, interpretation, and composition. Search engines weakened content memory but strengthened retrieval, filtering, and source evaluation. Platform algorithms weakened mastery of the real city but forced drivers and riders to learn to read the system, circumvent rules, and cultivate their own data portraits.

A person could still say, I understand this system, I command this set of rules, I have a set of experiences, I know what to do.

This might be the last remaining shred of security for modern people in a hyper-specialized society.

The complete artifact is gone, the complete process is gone, a stable environment is also gradually disappearing, but at least a person can still hold onto one thing: I can, I know, I have judged.

This is my ability.

But AI is beginning to erode this layer of security.

The Collapse of the Thinking Process

The first thing to fall is doubt.

Doubt is the first line of defense for cognitive subjectivity. A judgment truly belongs to me not because it was sent from my account, but because it was once questioned, resisted, and tested by me.

What the Wharton School's 'cognitive surrender' experiment revealed is precisely how this defensive line is preemptively shut down by AI. People don't choose to trust AI after sufficient comparison; more often, they stop questioning the moment AI provides a fluent, complete, and confident answer.

The most dangerous thing about AI is not that it thinks for you, but that it makes you feel like you have already thought.

The second thing to fall is integration friction.

In the age of search engines, people had already begun outsourcing content memory. But search still retained many learning frictions: you had to click on web pages, compare sources, identify bias, sift through material, and reorganize conflicting information.

These actions seemed inefficient, yet that is precisely where learning happens.

Shiri Melumad from the Wharton School and Jin Ho Yun from New Mexico State University conducted a large-scale study published in PNAS Nexus in 2025. They ran 7 experiments with a sample size of over 10,000 people. Participants learned about a topic using either LLM summaries or web search, and then wrote a piece of advice.

The result was that the LLM group reported shallower learning and wrote advice that was shorter, more generic, and contained fewer factual citations. Independent readers also rated the LLM group's advice as less useful and less credible.

This shows that the hassle of searching is itself a form of learning friction. Search retains chaos, chaos forces engagement, friction forces judgment, and inconsistency between sources forces integration.

The LLM compresses all this into a smooth answer. What it takes away is not just the hassle, but the very path by which learning embeds itself in the subject.

Your cognition is the synthesis of questioning, verifying, doubting, judging, and integrating.

After one no longer doubts, no longer integrates knowledge independently, and just seeks a solved answer, cognitive participation almost completely disappears.

A team led by Nataliya Kosmyna from the MIT Media Lab conducted a study titled "Your Brain on ChatGPT." This study no longer relied solely on questionnaires or behavioral observation but directly used electroencephalography (EEG) technology to peek into what happens inside the human brain when using AI.

The research team divided participants into three groups: a pure brainwriting group, a traditional search engine-assisted group, and an LLM (ChatGPT)-assisted group. When participants wrote an essay on a complex topic (like what happiness is), the EEG data showed that the LLM group had significantly weaker neural connectivity and brain activation patterns during the task, with neural connections reduced by up to 55% compared to the pure brain group.

More frighteningly, in a subsequent recall test, the LLM group showed extremely poor memory. A staggering 83% of participants couldn't even cite a sentence from the article they had just 'written,' and they lacked a basic sense of ownership over the content they produced.

Participants 'wrote' an article, but they struggled to quote their own sentences and felt a lack of ownership over the content.

This is precisely the decoupling of cognitive product from cognitive participation.

What follows is the consequence of abandoning the thinking process: the loss of the ability-formation process.

Ability formation is not the entirety of AI's alienation, but it is the last credential modern people have to resist alienation.

In a society continuously narrowed by division of labor, job roles, platforms, and toolchains, a person can still use 'I can' to confirm themselves. I can write, I can judge, I can debug, I can understand a system, I can find the cause in an error.

In 2026, researchers Judy Hanwen Shen and Alex Tamkin from the leading AI research institute Anthropic published an in-depth study on how AI affects programmers' skill formation. This study hits the greatest pain point of the modern workplace: does using AI for efficiency equate to an improvement in personal ability? They had 52 developers with Python experience but who had never used Trio learn an unfamiliar asynchronous library. The AI-assisted group didn't look bad after completing the task. But in a follow-up closed-book test, they scored an average of 17% lower, with the biggest gap being in debugging ability.

This shows that completing a task and forming an ability are not the same thing.

AI's problem is not that it makes it forever impossible for people to learn, but that it makes the act of learning seem increasingly unnecessary.

After AI, you can still possess results, even more results. More text, more plans, more summaries, more code, more judgments. But these results no longer necessarily mean you have undergone the corresponding formation process.

Defining the Alienation of AI

Only at this point can we truly name the new alienation brought by AI.

It is not a simple degradation of ability, nor an ordinary efficiency boost. It is the rupture between cognitive outcomes and the subject's process of formation.

People possess more and more 'their own' cognitive outputs, but they are less and less able to confirm themselves within these outputs.

Taylorism turned people into hands in a production process. AI may turn people into signatures on a cognitive process. It no longer needs the whole you; it needs your confirmation, your signature, your submission, your preference, your endorsement, and your assumption of responsibility.

This is the true historical position of AI-induced Cognitive Surrender.

It is not another labor-saving trick in the history of tool outsourcing, but the moment when alienation penetrates from the labor process into the thought-generation process itself.

For the first time, humanity possesses on a mass scale cognitive outcomes that 'look like its own thoughts,' yet finds it increasingly difficult to confirm whether these outcomes truly ever passed through themselves, changed them, and became a part of them.

A Solution That Is Hard to Form

To prevent this alienation from truly taking hold, in 2026, Xu and colleagues proposed a method in their paper titled "Cognitive Sovereignty Surrender": scaffolded cognitive friction.

A truly excellent AI system should deliberately create a reasonable epistemological tension in its interactions, even acting as a 'computational devil's advocate' to interrupt humans' subconscious reliance on intuition.

A good tool must generate appropriate resistance, forcing humans to retain the agency to ask questions, seek evidence, maintain doubt, and make the final judgment.

But will you use an AI that creates friction, or one that gets the job done faster? And which product will a company seeking unprecedented business growth choose?

For today's users, the answer seems self-evident.

AI's new alienation is not caused by AI alone. It requires three conditions to be met simultaneously.

Because the engine of alienation rumbles loudly and will not pause for a moment.

We seem to have stepped onto a path of despair.

As cognitive sovereignty is fully taken over by AI, humanity appears destined to become pets bred by algorithms, gradually losing all ability to face the real world in the tepid water of zero friction.

04

The Engine of Alienation

In my youth, when I imagined technology liberating humanity, the picture that came to mind was often one of ancient Greek philosophers strolling in leisure through the Academy of Athens: machines laboring in place of humans, who would then retreat to their studies to engage in pure thought, art, and public life.

But reality's gravity is entirely different. Technology does indeed save us the time and mental energy spent on old tasks, but this freed-up 'cognitive bandwidth' does not return to ordinary people in an unconditional form. On the contrary, almost the instant it is released, it is immediately reabsorbed by the new tasks, new systems, and new evaluation standards of a rapidly spinning modern society.

Machines increased production capacity, so workers were demanded to submit to a more precise factory rhythm. Computers increased processing speed, so white-collar workers were required to take on more communication, explanation, and coordination. Platforms increased matching efficiency, so laborers were forced to respond faster, adapt more precisely, and submit more thoroughly to the evaluation system. AI increases the speed of cognitive output, so people are demanded to submit more plans, make faster judgments, make fewer mistakes, and be online more continuously.

Technology could have liberated people, but a growth-obsessed society will not allow what is liberated to remain idle. Thus, every liberation is recoded as an obligation.

The experiments above can only tell us one thing: humans have a tendency toward cognitive surrender in the face of low-friction AI. But a tendency is not destiny. What truly turns this tendency into a social structure is the growth society itself.

Within the organizational logic of growth and efficiency maximization, a faster tool will not automatically liberate people; it will immediately be recoded into higher output, faster response, fewer errors, and more responsibility. AI is therefore not the sufficient cause of alienation, but rather the amplifier through which the growth engine penetrates the cognitive production process.

The real precondition for AI to bring about alienation is our tacit acceptance of the growth-ism doctrine—treating the 'maximization of efficiency' as the absolute essence of human existence.

The fanaticism for growth and efficiency is the true engine of alienation.

The Engine of Alienation Started Late

But this chase for efficiency is not the whole of history; it is instead a collective personality trained into modern society over a very short period of history.

Humans, of course, have always wanted to live better and have always sought labor-saving methods. But 'better' does not inherently equal 'faster,' and 'labor-saving' does not inherently mean reinvesting all the saved time back into production.

For the vast majority of historical periods, human societies were not organized around annual growth, continuous efficiency improvement, and limitless output.

People lived more within the contexts of livelihood, religion, land, family, festivals, honor, craftsmanship, community, and order. Those societies were not idyllic, nor were they free, but they at least demonstrate that making efficiency maximization one's highest destiny is not the eternal backdrop of human civilization.

Growth becoming a moral virtue is a very recent development. It is neither in line with human nature nor the outcome of a free market alone.

Nobel laureate in economics Joel Mokyr, in A Culture of Growth, explains why sustained growth occurred in early modern Europe. His answer is not that humans naturally love growth, but that a special knowledge culture formed in Europe between 1500 and 1700.

During the Enlightenment, useful knowledge was promoted, technological improvement was given dignity, the scientific community and the Republic of Letters accelerated the circulation of ideas, and people began to believe that the world could be continuously improved through knowledge.

A growth culture is not eternal; it has a clear historical origin.

Starting with Weber's The Protestant Ethic and the Spirit of Capitalism, he argued that modern people do not naturally see continuous labor and accumulation as a virtue.

It once required a religious ethic to sanctify secular vocations, to shape restraint, diligence, and reinvestment into a life discipline born from salvation anxiety. The first step of growth culture was turning 'more' into 'more righteous.'

And the economic anthropologist Karl Polanyi, in The Great Transformation, outright argued that making the 'self-regulating market' the supreme organizing principle of society is entirely against human nature.

His evidence was derived from anthropology. Exchange, trade, and markets have long existed in history, but they were mostly embedded within kinship, religion, status, reciprocity, redistribution, and community orders.

People exchanged goods, but not always to maximize profit; it could also be to maintain relationships, fulfill obligations, gain recognition, secure a livelihood, or maintain social equilibrium.

In modern times, labor, land, and money are transformed into commodities, and social relations are made, in turn, to serve the price mechanism. Thus, profitability is no longer just a motive in certain transactions but is elevated to the moral code and discipline of the entire society's functioning.

The pursuit of profit is not the eternal essence of human economic life; it is simply the market society disguising its own historical form as human nature.

The Protestant ethic sanctified secular professions, giving religious meaning to continuous labor, restrained consumption, and capital accumulation. Industrial capitalism pressed this ethic into the factory system, reorganizing human bodies with clocks, discipline, wages, and productivity. Taylorism further broke down labor into measurable, optimizable, and replaceable motions. By the 20th century, GDP, national economic accounts, post-war developmentalism, and Cold War competition turned the growth rate into a scorecard for nations to compare against each other.

And now, on this path of a growth society, humanity is destined to choose to replace its own judgment with AI.

But since this path was never originally in our nature, why is it the only solution?

What truly determines whether AI pushes people into deeper alienation or helps them escape some forms of it is which societal logic it is plugged into.

If it continues to be plugged into the growth engine, then it will only transform cognitive surrender into a new production discipline. It will make people produce more 'their own' texts, plans, and judgments faster, yet they will form themselves less and less within these outcomes.

In the Era Before Full AI Replacement, We Face the Most Exhausting Liquidity

Of course, the socio-cultural engine layered over centuries of history will not stall immediately. We are about to face a near-term future that is almost irreversibly bleak.

According to the weak link theory in economics, AI will not directly replace humans but will instead exacerbate the previous processes, relocating humans onto a narrower, higher-pressure, and more unstable production chain.

In a highly complementary production system, the overall output is not determined by the fastest link but is bottlenecked by the slowest, hardest-to-compress, hardest-to-automate link. AI can quickly make writing, programming, searching, drafting, summarizing, customer service, scheduling, and data analysis much cheaper. But as long as the entire chain still involves responsibility, trust, coordination, exception handling, client relationships, ethical judgment, organizational politics, and final sign-off, the system will still need humans.

Humans are compressed into a responsibility bottleneck. More low-intensity tasks will be siphoned away, and any breathing room will be eliminated. What remains is almost entirely high-density judgment. In this accelerating process of skill obsolescence, what you learn increasingly resembles a temporary pass, only ensuring you won't be swallowed by the next round of automation. A person is still in the flow, but can no longer be called a labor subject; they are more like a stamping point on a cognitive production line.

Fluidity, insecurity, and exhaustion will reach their peak in this unfinished, intensifying alienation.

Until one day, either this growth engine no longer holds the power to explain everything, or humans are completely excluded from the labor cycle.

Only then will another path appear. Only then might the cognitive bandwidth released by AI not be immediately captured by the performance system.

05

The Possibility of Another Path

Oxford philosopher Nick Bostrom, a seminal figure in current mainstream AI ethics theory, attempts to answer a question in his book Deep Utopia: When a super-AI perfectly solves all instrumental problems, and all human material needs and daily decisions are properly taken care of, what reason is left for humans to act?

In that era, the value of growth for an individual will necessarily be greatly diminished.

He envisions that perhaps, after that point, humanity will fully transition to 'Autotelic Activities.'

These activities are valuable not because they produce some useful outcome (since a machine could always produce it better), but because the process of the activity itself has intrinsic value, forming a self-sufficient loop.

A person is no longer a cog mechanically moving to complete a task, but a living subject existing for relationships, play, aesthetics, exploration, and self-cultivation.

Efficiency and ability may both become far less important. What a person must answer are their own questions about meaning.

So, if you ask me, 'How can you get the most benefit from AI?' My answer is: survive the present, and wait for the almost inevitable stalling of the engine of alienation.