Micah Zarin's Blog

Let us begin with a premise that, while still disputed by some, will eventually seem so obvious that arguing against it will appear quaint: artificial intelligence, if development continues without interruption, will be capable of producing almost any output a human can.

There is no reason to suspect that such a time (one where AI output is indistinguishable from human output) is far away. Personally, I suspect that we will reach this point in the year 2030. Then, there is not much stopping CEOs from automating most jobs. Of course, certain jobs that require an amount of interpersonal communication may not get automated. But, just because a job is not automated at a given moment, does not mean it is safe. Thus, my point is not that almost all jobs will be automated by 2030. Rather, I believe that almost all jobs will be “automatable.”

Anyways, if this is, as many predict, the future environment, then what skills should one obtain now? Or, is this a very distant future? Is there a point in even learning how to think about anything?

There are at least two responses to this. The first is that thinking, if treated as a means to an end: problem solving, analysis, decision-making, is increasingly inefficient compared to what AI can do. A machine with access to comprehensive data, probabilistic models, and recursive feedback loops will out-think the average human, and eventually, perhaps, the exceptional ones too. In this domain, “thinking” is best outsourced.

But the second response begins where the first leaves off. Thinking is not only functional. It is also formative. When a person thinks, they are not merely generating conclusions, they are undergoing a process of becoming. The act of wrestling with uncertainty, organizing internal conflict, distinguishing belief from assumption, engaging with contradiction; this is not efficient. But it is constitutive of the self.

That may sound romantic, but it is not. It is merely descriptive. Most of the time, we do not think to solve problems. We think because it is how we orient ourselves. But we also think about thinking. We must decide what things are worth thinking about in the first place. This skill of “meta-thinking” will be a deeply useful one in the future. We will see a shift from numerical thinking, and hand-coding, to prompting. Knowing what questions are worth asking an AI in the first place. For example, an AI can tell you various ways to make money. This is, in some sense “thinking.” Or rather, it is indistinguishable from the output of a human thinking. However, whether or not money is the metric you should use for your happiness is also a question of thinking. The difference is that the ladder question is a matter of meta-thinking (which, I think, is best taught in a philosophy degree, so I guess there is your answer).

But don’t click off yet. Much depends on what replaces economic compulsion. If income is decoupled from labor, via either universal basic income, shared ownership models, or resource abundance, then work becomes elective. But elective work is fundamentally different from required work. What remains is not work, strictly speaking, but activity. And not all activity sustains the self.

This is why meaning will emerge as as a daily problem. If you no longer need to work, and your needs are met, then you are left with a blank canvas. For some, this is terrifying. For others, exhilarating. For most, confusing.

And so, “meta-thinking” (i.e., philosophy) is equally relavant. Philosophy trains people in how to relate to questions that do not resolve. The problem of purpose, the question of value, the nature of the self, these become practical concerns when material scarcity recedes. A person with no job and a guaranteed income will eventually ask what they are doing, and why. Not once, but repeatedly. Philosophy is not the only way to explore these questions, but it is one of the few disciplines that refuses to resolve them prematurely.

It is also one of the few disciplines that resists optimization. AI may be able to generate philosophical arguments. It can already write essays on Plato and simulate debates between Kant and Deleuze. But it cannot experience confusion. It cannot feel the dissonance of holding incompatible beliefs, or the discomfort of living with partial understanding. And these, awkwardly, are where real philosophical insight begins.

We should not overstate this. Philosophy degrees are not a cure for economic displacement. Nor will every disoriented citizen of a post-work world enroll in graduate seminars on epistemology. But the habits of mind that philosophy cultivates, like reflection, interrogation, tolerance of ambiguity, may become socially valuable in a different way. They allow people to function in a world where capital is no longer the organizing principle.

This does not mean we will become a society of philosophers. More likely, we will become a society that needs what philosophy offers: orientation.

It is possible that most people, faced with the collapse of work and the redundancy of thinking-as-function, will turn toward distraction. This is already visible in the proliferation of content, gamified apps, and identity performance. But this is not new. Human beings have always oscillated between engagement and avoidance. The question is whether society offers alternatives: practices and roles that help people integrate freedom rather than be hollowed by it.

That is what’s at stake. Not whether AI will take over work (it will), or whether it will out-think us (in many domains, yes). But what we do when the old pressures are gone, and the new ones are internal, vague, and optional.