Like all of us who were alive and sentient on that clear blue Tuesday (maybe especially those of us in and around New York), my memories of September 11th are etched deeply, permanently in my heart.
I was eleven years old, just starting middle school. Still, I could tell the adults were just as afraid as I was. No one had lived through anything like this. In the suburbs of the city, fighter jets flew menacingly over our house. Mom and I huddled by the TV watching our beloved New York burning, unrecognizably, like a warzone. I slept in her bed that night, and for many nights after, neither of us wanting to be alone.
No one knew what was happening, or why. No one knew what would come next. And it felt distinctly possible we could die.
I had never felt that kind of existential fear, before or since—until this week, when Jacob Coxon resigned from Anthropic, saying (in part): “The people building AI earnestly believe that it could kill us all by the end of the decade,” making horribly plain the stakes of the breakneck sprint toward artificial superintelligence.
The steady drumbeat of AI news has (like much news since I had my first baby two years ago) largely marched by me until now. I don’t find tech interesting. Just as I never watch sci-fi movies, conjecture about whether AI would one day turn against us and kill all of humanity seemed to be just that: conjecture.
Clearly it isn’t. This week’s revelations injected me with cold fear: the same fear I felt as an eleven-year-old girl huddled close to my mom on September 11th. The more I read, the clearer it became that Coxon’s warning is well supported by evidence, and no one (even the really smart people) seems to know how to safeguard against the worst possible outcomes. This is all exacerbated by a ghoulish Prisoner’s Dilemma: even if we’re able to corral the federal government into regulating American AI companies to make AI safer, this might stall us enough to lose the AI “race”, leaving actors less concerned with safety owning the dominant AI models, and putting us all in greater danger.
Once again, everyone is scared. No one has seen anything like this before. And once again, it feels distinctly possible we could die.
On the evening of September 11, 2001, my mom and I drove to church. There was a small gathering of people there who, like us, didn’t know what else to do but be together, and pray. I remember the dark sanctuary, and a small number of us packed in close. I remember Mom and I sitting down next to Tom and his daughter Nicki, our friends. Nicki looked ashen and was clutching her cell phone. When our pastor asked if anyone had specific prayers they’d like to lift up, Tom—normally a sort of surly, stoic guy—stood. Barely able to hold back the emotion in his voice, he asked us all to pray for his son-in-law, Nicki’s husband, who worked in the second tower and whom they hadn’t been able to reach. Nicki was very pregnant. I’ll never forget the sound of Tom’s voice. I’ll never forget the way the cellphone looked in Nicki’s hand.
I think about that night a lot. What it felt like to be with other people who were scared, and grieving. To be in a quiet, sacred place—listening. To all be faced with the same terrifying threat and follow our human instinct to be together, and support each other.
We need to get to church now. The challenges AI presents are enormous and complex. There seems not to be a single villain to punish, or one easy solution to implement, as satisfying as either would be. The scale of this moment calls for us to think and coordinate globally in an unprecedented way, a way that feels impossible given the vicious polarization now defining every debate. Facing these challenges will demand hard thinking, courage, and humility. It will require sharing information with people we may not like, or trust. It will require being, at least for now, on a team with people (maybe with nations) we don’t consider teammates. Liberal tech skeptics like me will need to humble themselves enough to take these AI companies at their word when they say they are trying to win the AI race in large part to keep us safe. AI companies will need to humble themselves enough to engage honestly with regulation, and coordinate with competitors in a way that may dampen profits. It’s hard. We don’t have another choice.
I don’t want to die. I want to watch my sweet, rowdy, just-getting-started son grow up and experience every beautiful part of this complicated world. Maybe some portion of these tech guys are proceeding without care for who dies. But I don’t believe that’s a majority. I believe most of us want to live, and to see our children grow. If we share that longing, we may have just enough left in our collective spirit to figure this out.
In AI, we see another tower looming. We’re all scared, we’ve never seen anything like this before.
How will we respond this time?