We’ve been hearing from multiple people lately that there’s about a 10 percent chance that humanity will be extinct relatively soon because of artificial intelligence (AI).
And one of the popular and, for me, most perplexing responses to warnings like this? That Big Tech is exaggerating the danger to make itself richer and more powerful. The idea is that the handful of richest men who ever existed on this planet are so manipulative and eager for money that they are going to threaten us with death so that we give them even more power and they make even more money.
Wait…what?
Since When Is Killing Your Customers Good Marketing?
The argument is that portraying AI as extraordinarily dangerous will cause us to hand over even more power to them. This, in spite of the fact that many researchers over the years have verbalized the exact same argument with the exact opposite goal—that of motivating us to control Big Tech. It seems clear to me that acknowledging the potential destruction of humanity risks hurting Big Tech more than it could possibly help. But let’s consider the flip side of this argument, an alternative that IMO is more likely—that is, that it’s true that they are lying, and lying in a way that protects them, but it’s the opposite lie: They are minimizing the risk to humanity.
If anything, history gives us plenty of reasons to suspect that companies minimize dangers rather than exaggerate them; think Big Tobacco or Exxon. When enormous amounts of money are at stake, corporations have repeatedly downplayed risks to the general population. In fact, these Big Tech bros have already been found guilty in court of lying to us by minimizing the harms of AI, and owing consumers hundreds of millions to even billions of dollars. Yet, suddenly we trust them on the extinction bit?
It makes no sense to me that we rarely discuss the possibility that Big Tech is continuing to play by its playbook of understating the danger to humanity. Why assume these companies would suddenly change their approach on this one issue of extinction and tell us and their investors everything they really believe? To me, the much higher likelihood here is that they are minimizing the risk of human extinction.
Furthermore, we aren’t even talking about the possibility of civilization collapse. This is different from total extinction, and actually probably much more likely. Throughout history, human civilizations have collapsed, but some humans have lived on. Similarly, it seems that this possibility of civilization collapse would be even more likely than total extinction today. Most, but not all, humans dead? In fact, this issue was recently studied. A few months ago, MIT published a collaboration with the University of Queensland Psychology Department, surveying 272 AI specialists from 37 countries. The experts were asked to estimate the odds of a catastrophic outcome from AI in the next 5 years. Yes, you read that right: the next 5 years. (Catastrophic outcome was defined as more than one million deaths, more than $100 billion in financial losses, or comparable harms.) In sum, several risks were estimated at greater than 20 percent, while 18 of the 24 risks were estimated as at least 10 percent likely. Ultimately then, it may be more probable that we will experience the collapse of our civilization than total human extinction. So, if Big Tech is so motivated to be overestimating the risk of extinction, why are they literally ignoring the potentially even more likely possibility of civilization collapse?
I believe we owe it to our children, and our ancestors, to consider this.
Maybe We Just Don’t Want to Believe It
Humans are extraordinarily skilled at self-deception. As one obvious example, we insist we’re independent thinkers while readily accepting the influence of people literally called influencers, whom we know are financially motivated to manipulate us into buying stuff. Psychologists have spent decades studying why humans are so adept at avoiding uncomfortable realities, and examples are many. Cognitive mechanisms like normalcy bias lead us to assume that because things have always worked out before, they’ll continue to work out. Motivated reasoning supports our embracing explanations that allow us to feel safe and in control, even when we aren’t. Intellectualization lets us debate probabilities and motives rather than confront their implications. And when all else fails, there’s always the most basic of defense mechanisms: denial. My guess is that our tendency toward denial is exactly what Big Tech is counting on, and so far, we haven’t disappointed them.
We’ve Seen This Movie Before
Literally. It’s called Don’t Look Up. In the film, scientists discover a comet heading toward Earth. Yet humans go to great lengths to minimize or simply ignore the threat. The disturbing part of that movie isn’t the comet but how recognizable the human behavior is.
I’m coming to the conclusion that perhaps the most dangerous thing about AI isn’t that we don’t understand the warnings. It’s that we understand them perfectly well and have become exceptionally good at explaining them away. But denial isn’t inevitable, and if my years as a psychologist have taught me anything, it’s that denial generally leads nowhere good. Instead, owning our power and looking directly at what frightens us strikes me as our only realistic option. We can join together, demand answers, and insist that the future of humanity is far too important to leave in the hands of a few billionaires.
