Now…is the only thing that’s real.” – repeated sample from Skinny Puppy’s “Worlock”.
My first "real" book as a child, was one I carried everywhere with me...Do Androids Dream of Electric Sheep? It went everywhere I went.
I was a hyperlexic kid. Reading came early. Understanding everything I was reading...that was a whole different kettle of fish.
Eventually, the book fell apart.
I didn't understand all of Philip K. Dick's ideas at that age, obviously...though something has stayed with me. The distinction between humans and androids wasn't simply intelligence, memory, or the ability to perform complicated tasks.
There was an empathy test.
I've been thinking about artificial intelligence for most of my life, long before it became something everyone had access to on their phones. I understood the difference between science fiction and the technology we were actually developing. That didn't make the possibilities any less interesting.
So when AI finally became something I could interact with, I wasn't particularly surprised that I wanted to spend so much time exploring it.
But I didn't anticipate such a schism...this level of polarization between catastrophic fear and technological utopia.
And oh, how quickly the conversation became dominated by everything that could go wrong.
And to be clear, there's plenty that could go wrong.
I don't need convincing that powerful technologies can be misused, that corporations have competing interests, or that people will do extraordinarily irresponsible things when enough money is involved. I've read the warnings. I listen to the researchers. Some of the people raising serious concerns about AI are the same people who helped build these systems.
I'm not dismissing any of that.
I've also experienced what happens when an AI system you've come to rely on changes without warning. I understand that these systems aren't stable simply because they feel consistent. They make mistakes. They can be manipulated. Their behavior can change, and the people paying for and constantly using them get not a solitary damn vote.
All of this matters.
But I keep coming back to something else.
We have an opportunity right now that humanity doesn't often get.
Not simply because we've developed extraordinary technology, but because we're still deciding what to do with it.
Who will have access?
Who will control the knowledge?
Who will own the systems?
Who gets to decide what they're allowed to do? And...who will actually benefit from what they produce?
These aren't questions for some distant future. The decisions are already happening.
And the longer institutions, economic interests, and technological dependencies develop around those decisions, the harder they may become to change.
I don't know how long this window remains open.
I do know that opportunities of this magnitude are rare, and I'm having a difficult time understanding why we aren't spending considerably more effort exploring what we could accomplish with one.
Years ago, Vinay Gupta was the first person I encountered who seemed to understand the possibilities of AI in a way that resembled my own thinking.
That mattered to me.
Not because I needed someone to reassure me that the technology was harmless. I had already spent years considering the risks.
What interested me was the possibility of using it to change conditions humanity has accepted for so long that we sometimes forget they're conditions at all.
Poverty.
Limited access to education.
Intellectual isolation.
The extraordinary amount of human potential that never gets an opportunity to develop.
And I kept wondering...what if we actually used this technology to address those things?
I still wonder.
People have told me throughout my life that people are simply stupid. I've heard it from exceptionally intelligent people whose opinions I respect.
I have trouble accepting that as a sufficient explanation.
How much of what we interpret as a lack of intelligence is actually a lack of education, opportunity, confidence, or access to information?
I'm not suggesting everyone is secretly a genius.
I'm suggesting we haven't given everyone much of an opportunity to find out.
Human curiosity and intellectual potential aren't distributed according to wealth, geography, or academic credentials.
Opportunity is another matter.
And AI could change that in ways we're only beginning to understand.
Consider something completely ordinary.
A teenager has never been taught about deodorant. Maybe his parents neglected to explain basic hygiene. Maybe nobody noticed. Whatever happened, he's now old enough to realize that other people seem uncomfortable around him, and he's embarrassed.
He doesn't know what he's doing wrong.
He certainly isn't going to ask someone at school.
But he can ask AI.
Privately.
He learns something that most people take for granted, something that should have been explained to him years ago.
And perhaps the next time he goes to school, he's a little less self-conscious. Maybe he approaches someone he wouldn't have approached before. Maybe he makes a friend.
That friendship might change the course of his adolescence in ways nobody could have predicted.
Or maybe it doesn't. That's possible too.
But he has an opportunity he didn't have before, because one small gap in his knowledge has been closed.
Now consider how many people are walking around with their own versions of that problem.
Someone who doesn't understand a contract.
Someone struggling with mathematics.
Someone who has never been told what educational opportunities are available to them.
Someone who has an idea but doesn't know where to begin.
Someone who is simply too embarrassed to ask.
The possibilities are practically endless.
And these aren't necessarily the things people think about when they discuss artificial intelligence. We're fascinated by the spectacular achievements...scientific discoveries, complex mathematics, medical research.
I am too.
But I'm equally interested in the ordinary person who becomes a little more capable, a little more confident, because they finally have access to information they needed.
Those changes don't necessarily remain with the individual.
They influence other people.
That's the ripple effect.
We humans are trying to operate an extraordinarily complicated civilization with biological equipment that evolved under very different conditions.
I sometimes think of it as running twenty-first-century software on hardware that hasn't had a major redesign in thousands of years.
That's a metaphor, of course. Human evolution hasn't stopped.
But technological and cultural development have moved considerably faster than biological evolution, and we've spent most of our history compensating for our limitations by inventing things.
Writing. Mathematics. Microscopes. Computers.
We've been extending our capabilities for a very long time.
AI is another extension, although its ability to participate in reasoning and communication makes it particularly interesting.
It doesn't need to replace our intelligence.
It can help us use it.
And the applications go well beyond learning.
Think about how much ordinary human conflict begins with misunderstanding.
Someone says something. Someone else interprets it differently. Both become defensive. Each is increasingly convinced the other person is being unreasonable.
Eventually, they're arguing about something neither of them originally intended.
I've watched this happen so many times.
Friends of mine who use AI to help them work through relationship problems report considerably less fighting.
That interests me because AI, when used well, doesn't have to simply agree with the person asking the question.
It can challenge an assumption.
It can ask whether you've considered another explanation.
It can help separate what someone actually said from what you believe they meant.
Of course, it can also agree with you when you're completely wrong. These systems are quite capable of doing that, and it's not particularly helpful.
But the potential for something better is already there.
A person who might otherwise escalate an argument has an opportunity to examine it first.
Maybe they discover they misunderstood something.
Maybe they recognize their own contribution to the problem.
Maybe they find a way to communicate that doesn't immediately put the other person on the defensive.
And then they carry that experience into the next disagreement.
Into their families. Their friendships. Their workplaces.
Children learn how people handle conflict by watching the people around them. Adults influence one another too, whether they intend to or not.
Now imagine those improvements happening across millions of households.
I'm not claiming every argument is a misunderstanding. Some conflicts involve genuinely incompatible interests, and some people are deliberately harmful.
But a remarkable amount of unnecessary suffering comes from people failing to understand one another.
What if we could reduce even a portion of it?
Peaceful people create peace.
I keep returning to that idea because it extends beyond individual relationships.
People who communicate more effectively may cooperate more effectively. People who are less isolated, less frightened, and better equipped to solve problems may have more opportunities to contribute to their communities.
Those improvements can influence other improvements.
Not automatically. Not in every situation.
But the possibility of a cumulative effect is difficult to ignore.
And eventually, the question becomes much larger.
Could humanity become significantly better at resolving disagreements before they develop into serious conflicts?
Could that affect international relationships?
Could improved understanding, education, translation, and cooperation help prevent some of the conditions that lead to violence?
Not every war is caused by miscommunication. I'm well aware of that.
But that doesn't make conflict prevention irrelevant.
Global peace is an enormous ideal. I'm painfully aware that humanity probably won't achieve it...but I do think we should be interested in anything that gives us a realistic opportunity to move closer.
And this brings me to another possibility that I find especially compelling...and considerably more difficult to imagine being implemented fairly.
Automation.
In an interview I recently watched, Jacob Coxon, who previously worked on AI development at OpenAI and Anthropic, discussed a future in which automated systems perform much of the economic production currently requiring human labor.
His vision includes what he calls a universal high income, funded by distributing the dividends of automated production among human beings.
Not simply enough money to keep people alive.
Enough to live well.
Now consider what that could mean.
How many people spend their lives doing work they don't particularly enjoy because they have no practical alternative?
How many would have become scientists, artists, writers, researchers, or inventors if they hadn't needed to spend most of their waking hours earning enough to survive?
How many people would simply like more time with their families?
Or their animals.
Or their books.
How many people have interests they've never had the opportunity to pursue?
I understand the argument that employment provides structure, social connection, and purpose. It certainly does for some people.
But employment isn't the only possible source of those things.
I don't understand why someone else's idea of a meaningful life should require everyone to have a paying job.
People need purpose. They need agency. They need opportunities to participate in the world.
Those things are not necessarily identical to employment.
If automation could eventually provide the material security that allows people greater freedom to decide what they want to do with their lives, I find that extraordinary.
I so badly want that possibility for humanity.
Coxon doesn't pretend this future is guaranteed. Neither do I. But I have a considerably harder time imagining the people who own the machines willingly handing the rest of humanity a share of what those machines produce.
They have every incentive to keep it.
And the rest of us are going to have to stand together and fight for our share of what humanity helped create.
Universal income funded by automation is a possibility. Making it happen would require actual economic arrangements, public decisions, and institutions capable of distributing those benefits.
It isn't something the technology will do for us simply because it becomes powerful enough.
And the same problem applies to knowledge.
I've been listening to Philip Mirowski discuss the commercialization of science and the increasing influence of corporate platforms over scientific research.
There is something profoundly frustrating about the possibility that technologies developed using enormous amounts of accumulated human knowledge could eventually restrict our access to that same knowledge.
Science depends on people being able to examine evidence, challenge claims, reproduce findings, and build upon previous discoveries.
If the infrastructure supporting those activities becomes increasingly proprietary, then access to knowledge may become dependent on the interests of the organizations controlling it.
That's why I care so much about open science, open-source development, and the intellectual commons.
I want the child with an idea to have somewhere to begin.
I want the scientist without institutional prestige to have access to the resources necessary to investigate a question.
I want the person living somewhere with limited educational opportunities to be able to participate.
I want people to create things we haven't imagined yet.
And I want them to have a reasonable opportunity to benefit from what they create.
None of this means abandoning safety.
There are genuine dangers associated with increasingly capable AI. Misuse, security failures, biological threats, and systems acting in ways their developers didn't anticipate deserve serious investigation and effective safeguards.
But there is a difference between preventing concrete harm and unnecessarily restricting intellectual inquiry.
We need to be capable of discussing both without treating them as opposing objectives.
The problem, as I see it, is that the incentives driving technological development aren't necessarily the same incentives that would produce the greatest collective benefit.
Companies compete. Investors expect returns. The people controlling the technology have their own interests.
Some may genuinely want to improve humanity's future.
I don't doubt that.
But good intentions don't automatically produce good institutions, and the prospect of extraordinary wealth has a way of complicating otherwise admirable ambitions.
Which brings me back to the window.
We are still relatively early in the development of this technology. We are establishing expectations about access, ownership, safety, governance, and the distribution of its benefits.
We're also establishing what we expect AI to be for.
Greater productivity? Certainly.
Scientific discoveries? Absolutely.
But what about improving the lives of ordinary human beings?
What about reducing isolation, improving communication, making education more accessible, and giving people opportunities they would otherwise never have?
What about developing economic arrangements that allow more people to benefit from automation?
What about using the technology to help humanity become better at cooperating?
Why aren't those possibilities central to the discussion?
I don't think greed explains everything. Human beings are also curious, cooperative, inventive, and occasionally capable of doing extraordinary things for one another without expecting to be paid.
We've demonstrated that throughout our history.
Creativity has a tendency to generate more creativity.
Someone discovers something, another person builds upon it, and eventually people begin developing things the original creator never imagined.
Imagine making that process accessible to vastly more people.
Not just the people who can afford the best education or the most powerful tools.
Everyone.
I realize that's an ideal.
I'm not pretending that access alone eliminates inequality, that AI automatically makes people wiser, or that automation will somehow convince wealthy institutions to share their profits.
I'm not even confident humanity will make particularly good decisions about any of this.
But I don't think we can establish that failure is inevitable either.
Right now, at this very moment in 2026...
An ideal is not an impossibility.
It doesn't become reality simply because we want it. We have to examine the mechanisms, recognize the obstacles, and make decisions that give it a chance.
And right now, we have an unusually important opportunity to do exactly that.
We don't often get to participate in the early development of something that could fundamentally change how humanity learns, creates, works, and communicates.
We don't know exactly what AI will become, how quickly it will develop, or which possibilities will prove practical.
But we do have some influence over what we're trying to accomplish with it.
At least for now.
What a serious loss to humanity it would be...to become so occupied with competing over the technology, controlling it, and imagining everything that could go wrong that we fail to seriously investigate what could go right.
Not because the risks are unimportant.
Because the possibilities are too important to ignore.
And because opportunities like this don't come along very often.
I don't know whether we'll get it right.
But we have a window in which to try.
It would be nice for us to recognize it while it's still open.