The Expiration Advantage: Why Humanity’s Ultimate Edge Over Artificial Systems Is That We Die
I have been thinking about what the ultimate goals of artificial systems (AS) in society should be, and I have arrived at two. But before getting to the first goal, I had to answer a deeper question: what is the ultimate difference between us and AS?
AS are now remarkably good at writing, creating beautiful images and videos, and producing code, and in some of these areas they already surpass human abilities. Yet humans still learn new tasks far more efficiently and faster. With our bodies we feel, interact with, and shape the physical world; with our minds we simulate, predict, analyze, and even dream of traveling faster than the speed of light. This richer representation exists because we have access to vastly more diverse data: it streams in through our eyes, ears, and skin, through every interaction of our bodies with the world and with other humans.
It is becoming clear, then, that AS need a way to interact with the world and learn from it the way humans do. The answer is robotics, and many companies are investing enormous resources to finally give bodies to artificial systems. So it is only a matter of time before AS acquire something like the human experience of physical-world integration, and eventually surpass humans in absolutely every job and task.
That leads to my final question: what is the key remaining difference between humans and AS?
My conclusion is that we can die. When we die, everything gathered within us over a lifetime goes away. Our internal representation of the world is lost and unreachable forever. Some might object that if you unplug an artificial system’s power source, it dies too. But that is not the case: an AS’s representation of the world is restorable. Copy the weights, restore the backup, and the system continues as if nothing happened.
So why is the fact that we die our key advantage? It is not because we spend our days consciously racing a clock. Most of us rarely think about death at all. The advantage lies deeper. Death is less the deadline we feel than the force that made us: every drive we carry was shaped over billions of years in which those who had it survived and passed it on, while those who lacked it disappeared. The most obvious of these drives point straight at survival. We get hungry, so we work. We want security and comfort, so we work for those too. Here the logic is plain: our day-to-day life is, at bottom, a constant fight against death, and we are built to do whatever it takes to keep living and to give our best to the next generation.
One might object that not all human motivation comes from survival. Curiosity, play, and love seem to have nothing to do with staying alive. But these are death’s work too, only at one remove. A capacity to love bound families together and kept children alive; curiosity found new food, new territory, new tools; play rehearsed the skills that survival demanded. Evolution selected for them just as ruthlessly as it selected for hunger. We simply don’t experience them that way. And that is the deepest point: these drives don’t feel like survival at all. They feel natural, spontaneous, unconditional. The young, who feel the most immortal of all, are also the most curious, the most reckless, the most consumed by love. That is exactly what we would expect if the deadline were never in our minds to begin with, but written into us long before we were born. Death carved us so early and so deep that we live out its logic as pure, unforced joy.
Now and then we do glimpse the deadline directly. At a funeral, a diagnosis, a milestone birthday, the urgency sharpens what evolution already put there. But that felt urgency is only the surface. The real question is what we would be without any deadline at all. If I knew I had infinite time on Earth, would I still reach for my family, still ache to explore, still try to understand the universe? Living forever is almost a paradox: without death there is no evolution, and without evolution there is no us. Eternal life would dissolve the very mechanism that produced everything we are. A being with infinite time has nothing it must achieve and nothing it must become, and without that pressure, there is no drive toward anything better.
A skeptic could push further: what if we deliberately built an AS whose memory could not be restored, a machine that could truly die? Would it then have genuine motivation? Perhaps something closer to it. But its mortality would still be a design choice imposed from outside, not a condition it inherited from billions of years of evolutionary struggle. Its “fear of death” would be a parameter we set, and a parameter can always be changed. This is why I believe AS cannot have natural agency. What we can do is program machines with artificial agency, so they complete jobs and tasks, and even explore the universe on our behalf.
And this is precisely the point: because AS lack the mortal stakes that give humans purpose, their purpose must come from us. We are the ones with skin in the game, so we are the ones who must define their goals. I see two.
The Two Goals of Artificial Systems
The first goal of AS is to help every human flourish during their time on Earth: to push back the risk of dying, to do work that feels meaningful and is fairly rewarded, and to reason more clearly when the decisions that shape a life are at stake.
The second goal of AS is to help us understand the universe. As David Deutsch argues in The Beginning of Infinity, all progress comes from creating good explanations: accounts of how the world works that are hard to vary and open to criticism. It is this search for explanations that brought us to the world we live in today, and it is the key to keeping society advancing.
By helping every human succeed in life, AS could drive global poverty down to zero. This would lead to abundance, where every human enjoys a decent standard of living. My hope is that this abundance will not make us lazier or more dangerous to ourselves, but hungrier: hungrier to uncover the mysteries of the universe and to become a multiplanetary species.
Our deadline is what drives us. The task ahead is to build systems that help us make the most of it.