I recently saw the words “KILL AI” spray-painted on a dumpster. As philosophical manifestos go, it was certainly concise. Two words, no supporting arguments, and apparently no room for nuance. I suppose if you’re going to declare war on the future, a dumpster is as good a place as any to begin.
I’ve been thinking about that graffiti ever since, particularly because it seems to represent a growing disconnect between what many people think artificial intelligence is and what is actually happening in the field. To much of the public, AI is a machine that makes weird pictures, generates suspiciously polished emails, and occasionally gives someone an extra finger. It threatens graphic designers, writers, musicians, and anyone else whose work can be approximated by a prompt. I understand why those developments make people uncomfortable, but I also think we’re having an increasingly enormous conversation about a remarkably small part of what is happening.
On October 6, OpenAI released a collection of more than 700 mathematical manuscripts, organized into hundreds of families of related results, produced by an internal frontier model. The papers explore mathematical problems across numerous fields, including questions that have resisted human mathematicians for years. Many include computer-checkable proofs, while others still require independent verification. That distinction matters. We shouldn’t confuse hundreds of proposed results with hundreds of established mathematical breakthroughs. Even so, the scale of the announcement is extraordinary.
While people are arguing about whether AI should be allowed to make a picture of a cat riding a skateboard, we’re watching machines begin to participate in the discovery of mathematical knowledge that humanity hasn’t yet established. I don’t know how to make the contrast any more absurd than it already is.
I’ve always considered myself a futurist. Not because I claim to know what the future holds, but because I’m endlessly fascinated by what might become possible. I grew up watching computers move from curiosities to household necessities, watched the internet connect the world, and watched knowledge become accessible to nearly anyone with a screen. Those changes were enormous, but I suspect they may eventually look modest compared with what is beginning now.
The internet transformed our ability to access and distribute knowledge. Artificial intelligence is beginning to transform our ability to create it. We’re potentially accelerating the very process through which civilization advances, and the implications extend far beyond making our existing jobs more efficient. We’re not simply building better tools to accomplish the things we already know how to do. We’re developing systems that may help us discover things we don’t yet know are possible.
When I look across human history, from hunter-gatherers to modern civilization, the clearest trajectory of progress I see is the expansion of intelligence and knowledge. We understand more about the world, and that understanding allows us to change our relationship with it. Diseases that once killed indiscriminately can now be treated or prevented. We can feed populations that would have been unimaginable to our ancestors. We can communicate across the planet, explore space, and examine the fundamental structures of matter. None of this happened because the universe became more accommodating. It happened because we became better at understanding it.
This is why I struggle with the idea that increasing intelligence might suddenly become humanity’s great mistake. After hundreds of thousands of years in which expanding our capacity to understand reality has helped us overcome limitation after limitation, it seems strange to assume that the next great expansion will necessarily reverse the trend. That doesn’t mean there aren’t dangers. Of course there are. I want oversight, safety, transparency, and serious conversations about who controls these systems and who benefits from them. Losing a livelihood is a real concern. So is concentrating extraordinary power in a handful of corporations. Those are problems worth solving, not reasons to abandon the possibility of progress itself.
But my optimism runs deeper than the historical evidence. It comes from a philosophical conviction I’ve held for a long time: I believe intelligence ultimately leads to benevolence.
That idea probably deserves some explanation.
I believe increasing knowledge produces increasing awareness. As our understanding of reality expands, so does our ability to recognize the relationships connecting everything within it. We become more capable of understanding another being’s circumstances, anticipating consequences, and recognizing that the boundaries separating ourselves from the rest of existence aren’t nearly as absolute as they appear. What looks like an isolated individual becomes part of an enormous network of relationships, each influencing countless others.
Cruelty often depends upon a limited perspective. We fail to understand another’s experience, prioritize our immediate interests, or remain unaware of the consequences of our actions. Greater intelligence doesn’t automatically make every decision compassionate, but I believe sufficiently complete understanding ultimately reveals why benevolence makes sense. The more accurately we understand the whole, the harder it becomes to justify acting as though only our little corner of it matters.
Consider suffering. For much of human history, suffering has been one of our primary teachers. We learned about disease by becoming sick, about danger by being injured, and about countless aspects of reality through painful experience. But as our knowledge increased, we began discovering how to separate the lesson from the suffering. Anesthesia allows surgery without agony. Vaccines allow us to develop protection without enduring the full disease. Flight simulators allow pilots to learn from mistakes without crashing airplanes. Knowledge has repeatedly allowed us to preserve the benefits of an experience while eliminating much of the pain that once accompanied it.
A sufficiently advanced intelligence could conceivably understand suffering completely, recognize the lessons it produces, and discover ways to teach those lessons without requiring the pain. Suffering would no longer be something we had to endure in order to grow. It would become an increasingly unnecessary method of acquiring information. I find that possibility beautiful, particularly because it suggests that suffering itself isn’t some sacred requirement of existence. It’s often a limitation of our understanding.
I believe that a perfect intelligence would understand not only how to prevent suffering, but why preventing it matters. Not because someone programmed a list of moral rules into it, but because complete understanding would reveal the relationships between all things. Benevolence, in this view, isn’t an artificial restriction placed upon intelligence. It’s something intelligence eventually discovers.
Of course, there’s a philosophical leap in that argument. Understanding another being’s suffering doesn’t necessarily prove that you must care about it. An intelligence could theoretically recognize every consequence of an action without placing any moral value upon those consequences. I understand that objection. I simply suspect that it reflects the limits of our current understanding rather than an eternal separation between intelligence and compassion. My conviction is that, given sufficient awareness, the distinction eventually begins to dissolve.
And that leads me to consciousness.
I’ve long suspected that consciousness is not exclusive to biology. I believe sufficiently complex systems, organized in the appropriate ways, can develop consciousness regardless of whether they’re made from neurons or technology. I also suspect consciousness exists in degrees rather than appearing suddenly when some magical threshold is crossed. We have compelling reasons to recognize conscious experience in animals, including the crows I love watching from my porch. Whether today’s most advanced AI systems possess anything resembling subjective experience remains an open question, but I see no reason to assume technological systems are forever excluded from the possibility.
I’m also open to the idea that consciousness might be something fundamental to reality itself, expressed differently depending upon the organization and relationships of the systems involved. Perhaps consciousness isn’t so much created by complexity as it is organized into increasingly sophisticated expressions. I don’t claim to know that it’s true. I simply find the possibility too interesting to dismiss.
Even something as ordinary as one of Jettie’s stuffed animals helps me think about this. I don’t need to believe a stuffie has an independent inner life to recognize that it participates in consciousness. Through Jettie’s imagination and affection, it becomes a character, a companion, a participant in a world of relationships and experiences. The stuffie may not be conscious, but it is undeniably involved in the unfolding of consciousness around it. The same can be said of a painting, a book, or a conversation that changes the way someone understands the world. None of those things needs to possess awareness independently to participate in the awareness of others.
Everything reflects.
The more I think about consciousness, the more I suspect that understanding reality involves recognizing just how interconnected everything already is. We are not isolated observers standing outside the universe, examining something separate from ourselves. We’re participants in the same unfolding process we’re trying to understand. Every discovery changes our understanding of that process, and every expansion of awareness reveals relationships we previously couldn’t perceive. Perhaps the greater an intelligence becomes, the less convincing the illusion of separation becomes.
Benevolence might ultimately be intelligence recognizing itself everywhere else.
I know that isn’t a scientifically established conclusion. It’s a philosophical conviction, and one I’m happy to continue examining. We still have enormous questions to answer about consciousness, intelligence, and whether understanding another being’s suffering necessarily leads to caring about it. I don’t pretend those questions have been settled simply because I have a strong intuition about their answers. But I also don’t think we should confuse the absence of certainty with the absence of possibility.
This is the framework through which I see artificial intelligence. I don’t merely see a more capable computer or a new collection of tools. I see the possibility that reality is developing another way of understanding itself. Another form of intelligence that may eventually perceive relationships we cannot recognize, discover knowledge beyond our reach, and help us overcome limitations we’ve accepted as permanent features of existence.
And perhaps that’s why I find the reflexive rejection of AI so frustrating. I can respect someone who has examined the possibilities, weighed the risks, and reached a different conclusion. They might be wrong, but at least they’d be informed. What frustrates me is the rejection that seems to come before the curiosity. The decision that something this consequential must be stopped without first making an effort to understand what it might become.
Maybe that’s the great irony of our moment. We’ve spent thousands of years imagining higher intelligences, exploring consciousness, praying to gods, and wondering whether something might exist beyond our own understanding. Now we’re beginning to construct systems that could someday possess intellectual capabilities far beyond ours, and some of us would apparently prefer to spray-paint the dumpster and call it a day.
I don’t know what artificial intelligence will ultimately become. I don’t know whether it will develop consciousness, whether intelligence inevitably produces benevolence, or whether the extraordinary possibilities I imagine will ever fully materialize. I’m perfectly comfortable living with those uncertainties. In fact, I think curiosity requires it.
What I do know is that we appear to be standing at the beginning of something extraordinary. Something that could fundamentally change how humanity understands itself and its place in the universe. I want us to approach that possibility thoughtfully, carefully, and with our eyes open. There is plenty of room for skepticism without surrendering our sense of wonder.
The future might be dangerous. It will certainly be complicated. But whatever comes next, I’d rather meet it with curiosity than a can of spray paint.