AI Won't Kill Us All. Only We Can Do That.

Everywhere you looked last week there were headlines about how AI could kill us all. The catalyst was the resignation of Jacob Coxon from Anthropic and his public statement that his employer and OpenAI are racing toward self-improving superintelligence and gambling with our lives. The people building this technology, he wrote, earnestly believe it could kill us all by the end of the decade. Rather than squash the claim, Anthropic’s own alignment lead, Evan Hubinger, confirmed it publicly, putting his personal odds that AI wipes out humanity in the next decade above one in ten, while acknowledging the company has no plan yet for aligning a superintelligence.

Then, on a podcast that same week, I was asked what I thought AI is going to do to musicians and artists.

Those questions sit at opposite ends of the scale, one about the survival of the species and one about whether a songwriter can still make rent. But they share a hidden assumption, and it’s the assumption that’s doing the damage. Both are phrased as though we were describing the path of a storm. What will it do to us? How bad will it get? When does it arrive? The grammar concedes the outcome before anyone has made an argument.

We have started treating artificial intelligence as weather. Something on its own schedule, indifferent to our preferences, to be forecast and endured. It is nothing of the kind. It is a thing we built, and the role it plays in our lives is a thing we assign. Its power over us is the power we hand it.

Everything Around You Was Made Up

We forget this about nearly all of our institutions. The economy is not a natural phenomenon. Neither are markets, corporations, credit scores, national borders, the forty-hour week, or the idea that a person’s worth tracks their productivity. Every one of them is invented. We made them up, agreed to them, wrote them down, and then, within a generation or two, began speaking about them as though they were laws of physics we had merely discovered.

That forgetting is the most expensive habit our species has. Once a construct passes for nature, it stops being a choice and becomes a condition. You don’t hold a hurricane accountable. You don’t reform the tides. And so we end up in service to the very systems we built to serve us, describing our own inventions as forces we’re powerless against.

AI is the newest entry on that list, and it is being naturalized faster than anything before it. Two years into a serious public debate, we already discuss its effects the way we discuss demographics or the seasons. That is a choice about language, and language is where accountability either survives or quietly dies.

Which is why I’m not persuaded by the extinction framing. It isn’t that the risks are imaginary; some of them are real and worth serious work. It’s that “AI destroys humanity” locates the danger in the machine and leaves us as bystanders to our own story. If AI ever does us catastrophic harm, it will be because someone gave it that objective, or because we set a goal carelessly and never bounded what counted as success. Both are human acts. There is no version of this where the technology is the author.

Two Levers, and Only One of Them Is Yours

Here I want to be honest about something the tidy version of this argument tends to skip.

There are two ways we give AI its role. The first is creation: what gets built, what it’s trained on, what it’s optimized for, how fast it ships. That lever sits in very few hands. A handful of corporations and the people who fund them are making those decisions, and if you’re a working artist in Nashville or a small business owner in Connecticut, you have essentially no access to it. Pretending otherwise is condescending.

The second lever is consumption, and that one is entirely ours. What we use. What we pay for. What we press play on. What we accept as good enough. Every recommendation engine on earth is tuned to behavior that we supply, in aggregate, every day. The companies holding the first lever are almost wholly dependent on the second, because a technology nobody adopts is a very expensive science project.

So the accurate statement isn’t that we all chose this. It’s that a few people chose to build it and the rest of us decide, daily, whether it has a place. That is less satisfying than a clean verdict and far more useful, because it tells you exactly where your leverage is.

What AI Cannot Touch

Now the artist question, which turns out to be easier than people think.

Nothing AI will ever do can replace your creativity. Not your capacity for it, not your need for it. The impulse to make something that wasn’t there before, to take what’s inside you and give it a form other people can hold, is not a service the market provides to you. It’s part of what you are. A machine that generates a song does not reach into you and remove your reason to write one, any more than recorded music removed anyone’s reason to pick up a guitar.

So the anxious version of the question, will AI replace artists, is confused at the root. It treats art as output and artists as a supply chain. If that were true the threat would be total. But the making is the point. The expression of purpose through creation is a feature of the human experience, not a job category technology can delete.

What people are actually asking is a different question, and a fair one: what role will AI play in the creator economy? Who gets paid, who gets heard, what fills the playlists and the feeds and the theaters?

Same answer. Whatever role we give it.

The Reckoning Is Ours

If AI-generated music, film, and art displace work made by human beings, it will be because we chose to consume it. Not because a model got too good. Because audiences pressed play, listeners didn’t care, and buyers took the cheaper thing. Every executive who swaps a composer for a prompt is responding to a market that showed no resistance.

That’s the reckoning, and it belongs to us. Blaming AI for the collapse of creative work is like blaming the printing press for what got printed. It’s an evasion, and a disarming one, because a species convinced it is being acted upon stops acting.

I’d take the harder and more hopeful position. How we function as a society is our choice. We can decide that human authorship carries value worth paying for, and then pay for it. We can demand tools that extend what an artist can do rather than tools that stand in for having done it. But whatever the result, it will not be AI’s fault. This technology exists for exactly one reason: we created it and assigned it a function. Period. Full stop.

Change the Conversation

There’s a posture problem here too, and it may matter more than any single argument.

Right now nearly every public conversation about AI is defensive. We speculate, often in a frightened and faintly paranoid register, about what AI is going to do to us. Which jobs it takes. Which industries it hollows out. Whether it ends the species. Even the optimists are mostly arguing about the severity of a fate they’ve already conceded. We are pontificating about our own helplessness, at volume, in public.

That is the wrong conversation, and we should stop having it.

The conversation worth having is what we intend to do with this. How we use it to cure what has been incurable, to teach children who have no schools, to hand a working artist capability that used to require a label’s budget, to free millions of people from labor that consumes their lives and returns almost nothing. Not “what will AI do to us” but “what will we do with it, and for whose benefit.” That reframe isn’t optimism for its own sake. It restores the only fact that matters: the decisions are ours, and they have to be made deliberately rather than absorbed by default.

And notice where the fear narrative comes from. It is every bit as manufactured as the last arrangement we accepted without examining. Consider the bargain most people have been handed: give the bulk of your waking life to building someone else’s business, and if you’re disciplined and lucky, you may scrape together enough to find some peace in old age, well after your healthiest and most productive years are behind you. Nobody would design that if they were starting from what a human life is for. It was never designed. It accumulated, and then we mistook it for the nature of things.

The story that AI is an unstoppable force acting upon us is the same kind of story, arriving in real time. We can decline to repeat it.

The Objection Worth Taking Seriously

The strongest argument against me has teeth, so let me put it plainly. Constructs are hard to exit. Money is invented, and try living without it. Telling a session musician who just lost her income that society chose this can sound like telling her she chose the weather after all. Calling something a human choice doesn’t automatically make it an available one.

That’s right about the mechanics and wrong about the conclusion. Accountability spread across millions of people is still accountability. A choice made inside systems a few companies designed is not fate; it’s a governance problem and a culture problem, and those are the two kinds of problem human beings have always solved once they decided what was at stake was worth it. The difficulty of changing a construct is an argument for urgency, not surrender.

And the alternative framing gives us nothing. If AI’s role is inevitable, there’s no case to make, no standard to set, and nobody to hold responsible. I don’t believe that, and I don’t think you do either.

The question was never what AI will do to us. It’s who we intend to be while it’s here.

At Spotlight, we work with artists, musicians, and creative entrepreneurs on the premise that a person is not a unit of output. If this moment has you rethinking what your work is for, I keep a few slots open each week — no pitch, just a conversation. Book an introductory call.