I had the source wrong.

For a while, I had been carrying around the idea that the head of Anthropic said neurodivergent people might be unusually well equipped for the rise of AI. The remark was actually from Alex Karp, the CEO of Palantir, who wrote that "the neurally divergent" would disproportionately shape America's future.

That correction made the idea considerably more complicated, which is probably why I haven't been able to stop thinking about it.

Palantir is not a neutral messenger. The company represents something deeply troubling to many people, including me. Its technology has become entangled with military power, immigration enforcement, government surveillance, and the consolidation of enormous amounts of information about human beings. These aren't abstract concerns about what technology might someday do. They're questions about what technology is already being asked to do, who gets to direct it, and who is made vulnerable when they do. (ACLU, Associated Press)

A troubling messenger can still say something worth examining.

As a kid, I was described in seemingly contradictory ways. Other kids thought I was weird. Adults called me "mature for my age" and "an old soul." One made me feel outside; the other sounded like praise. Looking back, they may have been noticing the same thing. I was paying close attention, following uncommon interests, and making connections I couldn't always explain.

I was also the weird queer kid long before I had useful language for either part of that experience. Being called weird wasn't always affectionate, but the word eventually became less of a judgment and more of an explanation.

Later, discovering that I was neurodivergent gave me another way to understand it. It didn't turn every difficulty into a hidden advantage, and I'm wary of stories that portray neurodivergence as a kind of creative superpower. Different thinking does not automatically produce better thinking. It doesn't guarantee taste, wisdom, empathy, or moral clarity.

But it can make a person less attached to the obvious route.

That matters when the obvious routes are changing.

AI is being incorporated into nearly everything around us, often before most people have had time to decide how they feel about it. Every few weeks, another tool promises to transform how we write, design, search, code, learn, hire, diagnose, or communicate. The language surrounding it alternates between inevitability and apocalypse, with surprisingly little room between the two.

I keep returning to William Powers' Hamlet's BlackBerry, a book about what he calls the "conundrum of connectedness." Powers looks backward to show that new communication technologies have repeatedly arrived with a mixture of excitement and distress. The details change, but the human response has a familiar rhythm. A new medium appears. People embrace the possibilities. Other people warn that attention, memory, language, or civilization itself will never recover. Eventually, the technology becomes ordinary enough that we forget it was once destabilizing. (Google Books)

That history is comforting, but only up to a point.

The fact that technological panic has happened before does not mean the object of the panic is harmless. Familiarity is not innocence. The printing press, photography, television, the internet, and the smartphone all expanded human expression while also changing who held power, what could be distributed, and how quickly manipulation could travel.

AI continues that history, but it enters a political environment already strained by anti-intellectualism and anti-truth. Expertise is treated with suspicion. Evidence is negotiated according to affiliation. The institutions that might have helped us understand a profound technological shift are themselves fighting to maintain public legitimacy.

Some of the skepticism surrounding AI is loud, reflexive, and alarmist. Some of it is completely warranted.

There are legitimate questions about labor, authorship, energy use, surveillance, bias, consent, misinformation, and the concentration of power among a small number of companies. These questions should be addressed honestly. But even healthy skepticism has done little to slow the arms race. Governments and corporations are not waiting for a cultural consensus before moving forward.

Refusal is a valid personal decision. It is not, by itself, a strategy for shaping what happens next.

Historically, artists and entrepreneurs have often been among the people willing to enter the uncertainty of a new medium. They experiment before its conventions have settled. They misuse tools in productive ways, expose their limitations, and show the rest of us what the technology might mean beyond its original purpose.

That doesn't make them evangelists. Curiosity is not capitulation.

Artists can work with a technology while criticizing the systems surrounding it. Designers can recognize a tool's expressive potential while asking who benefits from its adoption. We can be excited by what becomes possible and disturbed by the conditions under which it becomes possible. In fact, holding those reactions together may be more useful than choosing one and pretending the other doesn't exist.

I felt that tension while rebuilding my portfolio.

Claude helped me code components that would have taken considerably longer to make on my own, assuming I could have made them at all. Seeing an idea become interactive so quickly was genuinely thrilling. It didn't feel like asking a machine to be creative for me. It felt like gaining access to a material I hadn't been able to manipulate with the same fluency before.

But speed introduced a different problem.

When producing something becomes easier, producing the right thing doesn't.

The tool could give me options. It could help solve implementation problems, generate variations, and keep going long after I would have become tired. It could not decide which possibility felt alive. It could not tell me when something was technically impressive but emotionally empty. It could not know whether a strange idea was worth protecting or merely strange.

That was still my responsibility.

This is where taste becomes more important, not less.

Taste is sometimes treated as surface preference, a collection of references or an ability to identify what looks current. I think it is closer to judgment under conditions of abundance. It is knowing what deserves attention when almost anything can be made. It is recognizing the difference between an idea that is finished and one that has merely become polished. It is noticing when a plausible answer is dead inside.

AI can produce more, faster. It can make competence abundant. It can imitate established patterns with extraordinary fluency, and it can help people cross technical boundaries that once kept their ideas trapped in their heads. Even Anthropic describes its creative tools as a way to expand what people can attempt while acknowledging that the technology cannot replace human taste or imagination. (Anthropic)

If that is where we're heading, creativity will depend less on proving that we can manufacture an artifact and more on deciding why that artifact should exist.

That is not exclusively a neurodivergent ability. Nor do I believe unconventional minds will somehow inherit the future while everyone else is automated away. That kind of prediction is too clean for what is actually happening.

But people who have spent their lives looking from the side may have something useful to contribute. We're familiar with the possibility that the accepted way of seeing a problem is incomplete. We know that apparent irrelevance can become a connection, and that following an unusual interest sometimes leads somewhere important. We may also be a little less frightened by the moment when the existing rules stop making sense, because some of those rules never made much sense to us in the first place.

The value isn't weirdness by itself. It is weirdness joined with judgment.

Palantir's place in this story remains unresolved for me, and I think it should. A company whose work demonstrates many of the dangers of technological power can also be led by someone who recognizes the potential of unconventional cognition. One observation does not absolve the other. If anything, the contradiction is a reminder that unusual thinking has no predetermined moral direction. It can challenge power, or it can become very good at serving it.

Taste must include deciding what we refuse to make.

So perhaps the creative task in the AI era is neither resistance nor surrender. It is participation with friction. It is using these tools without allowing their assumptions to become our own, questioning the systems around them while remaining curious about what they make possible, and insisting that efficiency is never the only measure of progress.

Artists don't need to sell the public on AI. The companies building it have more than enough money to do that themselves. What artists can do is stranger and more necessary. We can expose what the technology flattens, find uses its makers didn't anticipate, and give form to questions that cannot be answered by another product demonstration.

We can help make the unfamiliar understandable without pretending it is safe.

We can be skeptical without becoming incurious.

We can enter the future without agreeing to all of its terms.

It's all very weird.

I enjoy weird.