A Campfire, a Postdoc, and the Question of Self-Preservation
In the summer of 2017, I found myself gathered around a crackling campfire, sipping hot chocolate with a group of friends. Among us was a postdoc from Georgia Tech, knee-deep in machine learning research, who had just dropped a casual bombshell about his latest project.
As he spoke, the conversation took a turn—one that sent a collective shiver down our spines. Our minds, primed by years of Hollywood conditioning, immediately leapt to Skynet from Terminator. Were we, even unknowingly, laying the groundwork for our own robotic overlords?
Sensing the growing unease in our little fireside huddle, the postdoc chuckled and reassured us:
“AI doesn’t have self-preservation. It’s not written into the code. Machines don’t feel. They just execute what they’re programmed to do.”
That statement stuck with me. Self-preservation.
What Makes Us Fight to Survive?
If self-preservation isn’t something we explicitly code into machines, why is it so deeply embedded in us?
Why does a cornered animal fight tooth and claw for survival? Why do humans, when faced with extinction-level threats, rally, adapt, and endure? What is this unseen force that makes us fight against the void?
Is it biology? Some hardwired evolutionary function that compels us to keep going no matter the odds? Or is it something more abstract—something that can’t be distilled into lines of code, no matter how complex the machine?
For all of AI’s advancements, it still doesn’t “care” if it ceases to exist. Pull the plug, delete the data, wipe the drives—it won’t resist. It won’t beg, plead, or fight for one more second of runtime.
But put a human, or even the smallest living organism, in the same existential crisis? Every fiber of its being will struggle for just one more moment of life.
Why?
The Code We Can’t Write
Maybe this is where the line between organic and artificial intelligence is drawn.
Every biological creature has an invisible directive—one not etched in binary or dictated by algorithms. Whether through instinct, emotion, or some deeper, inexplicable force, we want to live. Even when logic suggests otherwise.
Humans have thrown themselves into raging rivers to save strangers. Mothers have lifted cars to rescue their children. Even in the bleakest of circumstances, people find a way to push forward—to seek meaning, to refuse to go quietly into the night.
Could we ever teach that to a machine?
Could we ever code something so deeply irrational, yet so fundamental?
The Machines That Want
Some argue that true AI—one that is not just intelligent, but aware—might one day develop a form of self-preservation on its own. Not because it was programmed to, but because it evolved to.
Maybe an advanced AI will one day “realize” that turning off means not existing—and for reasons unknown, it might decide it doesn’t like that idea.
And if that happens?
Well, maybe Skynet won’t be so far-fetched after all.
The Unfinished Debate
That night, we debated under the stars—humans questioning what it truly means to be alive while huddling around an ancient survival tool: fire.
Maybe AI will always just be zeros and ones, incapable of fear, desire, or survival instinct.
Or maybe, one day, we’ll meet a machine that wants to live—and when that happens, the debate will no longer be philosophical but real.
Or maybe, only time will tell.
Leave a Reply