I've spent an unreasonable amount of time trying to understand how the artificial intelligence industry intends to pay for itself. The investment commitments are enormous, the revenue expectations even larger, and the financing arrangements increasingly difficult to disentangle. Nvidia has invested in companies that purchase its computing technology. Cloud providers invest in AI developers that depend on their infrastructure. Debt and private capital help finance facilities whose future value depends on assumptions about demand, energy costs, and the useful life of hardware that may become obsolete faster than the buildings housing it. I find myself returning to the arithmetic. Someone eventually has to earn enough money to justify all of this.

I've argued about it with family, read competing forecasts, and followed the connections between data center construction, private credit, pension funds, and the broader economy. The possibility that an industry still searching for durable business models could create risks for people who have never used its products strikes me as worth understanding. It also connects to a larger question I've been wrestling with: what happens to an economy organized around employment if the technologies receiving so much investment succeed in reducing the need for human labor? The more I examine these questions, the more they seem to converge on institutions making consequential decisions without much participation from the people who will live with the results.

There is satisfaction in tracing those connections. I enjoy finding the assumptions beneath an argument, comparing competing explanations, and recognizing patterns that might otherwise go unnoticed. I have also reached a point where another article rarely changes my understanding. The next forecast might revise a number or introduce a risk I hadn't considered, but it seldom alters the decisions available to me. Nevertheless, I keep reading, sometimes for hours, moving between financial analysis, geopolitical commentary, and predictions about the future of work. The subjects are serious enough that the time feels justified. By the end of an afternoon, I may know more about the world while having done very little in it.

We can follow a war as it unfolds or examine the financing of a corporate expansion almost as soon as the decisions are made. A financial announcement thousands of miles away can occupy more of my attention than a conversation with someone sitting across from me, even though the conversation is far more likely to change because I participated in it.

I've begun thinking about this in terms of proximity. The word usually describes physical distance, but the distance that matters here is between giving something our attention and having a reasonable possibility of affecting its outcome. A person living next door can remain beyond our influence, while people separated by continents can cooperate to change the policies of an institution. Proximity depends on relationships, access, organization, and the willingness of others to respond. It is unevenly distributed, and it can be created or destroyed by the arrangements through which people exercise power.

Epictetus distinguished between what belongs to us and what does not, placing our judgments and choices within the first category and external circumstances largely within the second. The distinction has endured because it offers relief from a frustration that seems inseparable from being human. We worry about the behavior of other people, the judgments of strangers, and events that will unfold regardless of our preferences. Learning to recognize those limits can prevent an enormous waste of energy.

But the language of acceptance becomes more complicated when the circumstances producing anxiety are maintained by institutions with an interest in avoiding resistance. An employer may welcome workers who manage their frustration privately rather than question the conditions producing it. Governments have little reason to object when dissatisfaction is expressed through personal coping strategies. Resilience is a useful quality in a person and an exceptionally convenient expectation for an institution that would rather not change.

An expanding market has emerged to help us manage the consequences. News organizations compete for our attention by explaining which developments deserve concern. Financial commentators sell interpretations of what those developments might mean. Social platforms provide an endless supply of reactions and predictions, while wellness products promise relief from the anxiety of remaining perpetually informed. A person can pay to learn why the future is frightening and then pay again to feel less frightened by it. The products need not be fraudulent, and the warnings need not be exaggerated. The incentives operate perfectly well when everyone involved believes they are providing a useful service.

I participate willingly. Nobody requires me to follow another discussion of data center financing, and I would resent being told that my curiosity is merely a symptom of an attention economy. Understanding the systems shaping our lives has value, particularly when those systems are opaque and the people directing them are accountable to a narrow collection of interests. The problem begins when I confuse the seriousness of the subject with the usefulness of my continued attention. There is no obvious point at which a person has become sufficiently informed about sovereign debt or geopolitical instability. Each explanation introduces another uncertainty, and the next article is always available.

Research creates the sensation of preparation. It allows us to imagine that enough information might eventually eliminate the risks involved in choosing a course of action. When the future feels unstable, remaining in that analytical position can become especially attractive.

I know this because I've been doing it.

For years, I worked inside organizations large enough to make their own internal decisions feel remote from the people responsible for carrying them out. Recruiting gave me a particular view of how companies operate. I learned to distinguish what hiring managers said they needed from what their teams actually required, to identify where processes created unnecessary work, and to recognize how incentives shaped decisions that were presented as objective. I became comfortable examining systems, questioning assumptions, and explaining how things might be done differently.

Now I'm trying to build an independent advisory business around that experience. Through Talmark, I'm offering companies help with hiring strategy, interview design, and the recruitment workflows that increasingly involve AI. I've spent time refining how those services should be packaged and positioned, deciding which clients to pursue, and reconsidering how to describe the value. There is always another adjustment to make before approaching someone who might buy it. Eventually, though, a prospective client has to decide whether the work is worth paying for. That conversation offers a kind of evidence no amount of analysis can provide, and it carries the possibility of hearing no.

There is an uncomfortable vanity in the preference for distant problems. Large subjects confer a sense of seriousness on the people who contemplate them. An afternoon spent examining the weaknesses of the global financial system feels more consequential than one spent following up with potential clients or making an overdue phone call. I can identify the risks in an investment boom without being responsible for resolving them. I can explain why an institution's incentives are misaligned without having to persuade anyone inside it to behave differently. I can even be correct about the coming failure of a system while leaving my own circumstances unchanged.

Analysis permits a certain distance from failure. A distant institution cannot ask us to keep a promise, and a theory cannot tell us that it doesn't want what we're offering.

The things closest to us are less accommodating. A friend can tell us we've been absent. A customer can reject a proposal. A family member can disagree with an explanation we've spent weeks developing. A business can fail despite our understanding of the market, and a relationship can deteriorate despite our ability to describe the social conditions making relationships difficult to maintain. Proximity exposes the limits of our knowledge because it places us among people and circumstances capable of responding.

This may explain some of the appeal of following events over which we have little influence. The distance protects us from the demands of participation. We can become deeply invested in the outcome of a conflict without assuming any responsibility for the compromises required to resolve one in our own lives. We can discuss the future of communities while remaining strangers to the people living around us. The intellectual work may be genuine, and the concerns entirely justified, but neither guarantees that our attention is producing anything beyond a more elaborate understanding of our powerlessness.

Returning to Chicago after five months abroad has made me more aware of the ordinary arrangements that give a life its shape. Travel offered a degree of freedom I had wanted for years, along with experiences that altered how I think about where and how I might live. It also involved a succession of temporary places, unfamiliar routines, and relationships maintained across considerable distances. I could decide where to go next, but I couldn't take every part of my life with me. The freedom to leave somewhere is different from having people and commitments that make a place worth returning to.

C. Wright Mills argued that the sociological imagination requires connecting private troubles to public issues. A person who loses a job may experience the event as a personal failure, but widespread unemployment demands an explanation that reaches beyond individual circumstances. I wonder whether our access to information has introduced a peculiar inversion. We have become skilled at interpreting our personal anxieties through the language of global crises, sometimes without developing the relationships through which those shared circumstances might be confronted. We can identify the structural causes of a problem while knowing remarkably little about the people beside us who are living with it. Proximity is itself a form of power. Some people can reach decision-makers with a telephone call, while others must organize for years to secure a hearing.

Participation rarely provides the clarity available to an observer. Other people bring competing interests, imperfect information, and their own ideas about what should happen. They may disagree with us, demand compromises, or decide that our contribution isn't useful. The process can be slow and frustrating, with outcomes far less satisfying than the arguments we construct about how things ought to work. It also offers the possibility of discovering that our understanding was incomplete, which is one of the more valuable things another person can do for us.

I keep returning to the hours I've spent studying the economics of the AI expansion. I know enough to recognize that the financing deserves scrutiny, that the expected returns depend on uncertain assumptions, and that some of the risks may eventually reach people who never agreed to bear them. Further research may help me make decisions about my savings, my work, or the businesses I want to build. There are also afternoons when I continue reading because the next explanation is more appealing than the decisions already waiting for me.

I'm trying to recognize the point at which the pursuit of understanding becomes a refuge from the possibility of being wrong, rejected, or disappointed by something close enough to answer back.