AI isn’t running water

It's something different

It’s a trope now: characterizing artificial intelligence by way of analogy to old, familiar, innovations. I was at a conference last week where this framing was on almost every presenter’s lips. It’s easy to see why: like steam power, electricity, and modern plumbing, AI appears poised to become an ever-present force, woven into every corner of our lives. Many commentators have described AI as a utility: “It’s the new electricity,” they say. “We’ll soon have AI as easily as we have running water, just turn on the tap.”

For a long time, I said this too. After all, general-purpose technologies typically have a profound, widespread impact. Electricity let us banish darkness and automate factories. Running water let us build clean, modern cities. It’s tempting to think that AI will do the same: quietly underpin our daily routines, from translating languages in real time to curating our social feeds.

But the more I see AI advance, the more I realize that the “running water” analogy misses something crucial. It fails to capture the profound sense in which AI is different from any of these previous technological leaps. When we talk about electricity or plumbing, we’re talking about physical infrastructure. These tools changed how we live by saving labor, improving sanitation, and boosting productivity. By contrast, AI speaks to something deeper: our understanding of what it means to be human.

For centuries, at least in the Western tradition, intelligence has stood as one of the pillars of our sense of self. We have long prided ourselves on our capacity for reason, problem-solving, and creativity. From the ancient Greeks to the Renaissance, from Descartes’ “I think, therefore I am” to the Enlightenment, we have placed rational intelligence on a pedestal, treating it as the defining trait that sets humanity apart from the rest of nature.

AI challenges that assumption directly. Now we see machines writing poetry, defeating humans in complex strategy games, and writing economics papers. We see that what we previously considered quintessentially human — factual knowledge, artful composition, logical analysis — can be performed by non-biological systems. These systems may not have consciousness, at least not in any way we currently understand. But they execute “intelligence tasks” with a speed and breadth that most of us will never match. And they get more intelligent with each new release.

This is what sets the AI transformation apart from all previous technological revolutions. The introduction of the water pump did not shake our core identity. The harnessing of electricity did not make us question the nature of humanity. Neither mass production nor the digital revolution told us that one of our most celebrated qualities, our brainpower, might be less unique than we’d believed.

This shift is not just about mimicking certain human abilities; it’s about the underlying principle that intelligence is just a matter of applying the right amount of compute in the right way. One of the fundamental lessons from AI research, articulated beautifully by recent Turing award winner Rich Sutton in his “bitter lesson,” is that compute power matters more than anything else.

If more computation leads to more “intelligence,” then it suggests there is nothing magical about the human brain. We, too, might be just the sum of computational processes running on a biological substrate. That thought unsettles us because it chafes against millennia of human exceptionalism.

Of course, it’s worth acknowledging that AI’s potential to upend our sense of identity doesn’t negate its practical wonders. Like electricity, AI really is shaping up to be a powerful, widespread resource. It helps doctors identify cancers more accurately, assists geologists in forecasting weather extremes, and can design more efficient supply chains. Over time, it could become a companion in every routine task we do. Yet even as we plug these systems into our workflows, an undercurrent of existential angst seeps in. What are we for, then, if an algorithm can do our jobs better? If creativity can be convincingly simulated¹ by a machine, what does that say about the human spirit?

Skeptics may argue that our response to AI is simply the latest iteration of the Luddite panic; that we always greet transformative technology with dread, then adapt and integrate. The printing press, railways, and computers all sparked outcries that society and the soul would be harmed. And indeed, in many ways, we adapted just fine.

But there is still a difference. Steam engines didn’t challenge our claim to unique intelligence. Telegraph wires didn’t ask us to rethink the nature of human thought. Electricity didn’t insinuate that it might rival or surpass our creative spark. The changes AI brings are not limited to external upgrades in productivity or convenience; they reach inside, prodding us to question the crux of human identity.

That is why I now find the “running water” analogy so incomplete. To be sure, AI is pervasive and essential. But it is not just another layer of infrastructure humming in the background. In the coming years, it will continue to amaze us with feats of reasoning and creativity. It will push us to confront, and perhaps even redefine, our relationship to intelligence and self-knowledge. It may force us to admit that our minds, extraordinary as they are, are not so different from sophisticated computation.

None of this means that our sense of self is doomed, nor that we must abandon the humanities in favour of code. If anything, this disruption could become a clarion call for more robust inquiry into human flourishing and identify beyond the anthropocentric intelligence paradigm. But even if we remain special in some respects, the scale and scope of AI’s challenge are unlike anything we’ve faced before.

We’ve long used technology to amplify our abilities: to dig deeper, run faster, store more data. But now we’re building machines that can think, and that cuts to the heart of who we are. For better or worse, that distinction sets AI on a different plane than plumbing or electricity. And it’s in that difference that our greatest societal, existential, and imaginative challenges lie.

1. and, one must ask, when it comes to the domain creativity, what is the difference between “simulating” creativity and being creative? Can one tell the difference in the output?