Peak Robot
Peak oil, artificial general intelligence, and the self-destruction we keep choosing. By Aastha Uprety
The term “peak oil” was first introduced in 1956 by Shell geologist M. King Hubbert. It described the idea that, since oil is a finite resource, we will one day reach the zenith of oil production, and then, slowly but surely, we will run out of oil. Crisis was to ensue. The idea scared us, but did not slow us down. Instead we kept digging, as if peak oil was a prize to win, not a thing to fear.
But it was neither. The calamity of peak oil has not happened, and we’re not as certain as we used to feel that it will happen anytime soon. We’ve been extracting the black gold from the earth more and more, and still more remains.
Most scientists and policymakers now believe we’re more likely to reach a peak in demand before supply as sun- and wind-powered energy becomes cheaper. The combustion cars that make up global traffic jams will be swapped out for electric vehicles, like a landbound ship of Theseus. Many years ago, experts didn’t expect clean energy to become so feasible so fast. We now know that soon we won’t need oil anymore; oil will need us. But we haven’t reached that elusive peak demand, either. And unfortunately, oil’s harms have already been felt.
Yesterday’s peak oil mania is today’s AI hype. “Artificial general intelligence” is a fantasy touted as an inevitability. It refers to the fact that one day, we will have created an AI that is of superior intelligence to humans. Some techie factions even believe that it’s their duty—their dharma, you could say—to help build that AGI, lest it come into existence and seek revenge against those who did not contribute to its development.
They fear it, but that fear propels them to realize it.
Journalist Brian Merchant, as well as other technology writers, have suggested that AI companies actually benefit from mongering fear about the apocalyptic sci-fi potential of their product. As the foremost experts, only they possess the knowledge to prevent AI from becoming evil, and as the oracles delivering the prophecy, they seal its inevitability. What follows must be unstoppable growth. We see it in the insatiable demand for data centers—hulking buildings computing and processing and sweating for every click of a button online, hundreds of thousands of gallons of water coursing through their skeletons to keep the machines from overheating—a feat of energy in service of down paying a karmic debt that may never come.
The looming threat of peak oil could have opened a path to an alternative energy future, one where we rely on more sustainable resources than just those that are finite. But to our demise, we decided oil should still be part of the equation—at least until it literally could not—and the fact that it was destructive to the earth was irrelevant. Peak oil was never a milestone we should have accepted, and certainly not one to have sought. Warnings from as early as the 1950s could have been heeded, inconvenient truths could have been confronted, and the black gold in the ground could have stayed there, untouched. We braced ourselves for impact from the hypothetical crisis of running out of fuel; all the while, the real crisis of darkening skies and boiling waters steadily worsened, unimpeded by the worry that the reason for its being could run out.
We’re at a similar crossroads now, and I fear we’re not taking a moment to pause and think.
I don’t want an online chatbot to translate my thoughts into essays. I like to do the careful work of selecting words, as if with a tweezer, and crafting sentences, as if arranging tiny gems into a delicate pattern. But powerful forces have decided that the obsolescence of that painstaking effort is an inevitable outcome. Apparently, whether or not it is desirable is not relevant to its fruition.
I don’t want to live in a city where the sky turns dusty orange for reasons hundreds of miles away or where bottom-floor apartments are a permanent liability—all thanks to a natural substance that, when burned, burns the rest of us with it. But we missed our chance decades ago to spurn that outcome, to keep it buried in the earth in favor of energy sources that wouldn’t destabilize the planet, ecosystems and economies alike. Peak oil was an invention that could have served as a warning but instead became a self-fulfilling prophecy. With its utterance into the world came an opportunity, a lifeline. Is that really where we want to go? In accepting the premise, we initiated the journey of preparing for it. But we had asked the wrong question. “Will we run out of this resource?” left untouched the vastness of “What will using this resource do to us?” And so we embarked on a path of collective hallucination: Well, it was going to run out anyway.
Choosing to believe that a far-off computer intelligence will one day take over the world has spawned a race to control AI before it gets to control us. Meanwhile, the present, dire, and genuine threats of AI to climate, labor, and human creativity are in plain sight. It didn’t matter then that oil could have run out—the real harm of burning it to its fullest, which we steadfastly ignored, was its capacity to destroy the earth. And it doesn’t matter now that robots could take over the world. Their devotees are already trying.
Aastha Uprety is a writer, editor, and lifelong student of political economy living in Queens, New York.
