Simon Clark

A man who builds artificial intelligence for a living said this week that there's better than a one-in-ten chance it kills every human being on earth within ten years. I read it before sunrise. Then the house woke up. Charlotte came downstairs looking for her shoes. Jack wanted to know what was for breakfast. I hugged them both and kissed them goodbye and Allison took them to school, like it was any other Wednesday morning in Greensboro. The man is a safety researcher at Anthropic, one of the companies actually building this technology. Not a fringe voice. Someone whose entire job is to look for the danger. He said his company is trying. He said nobody has solved the problem, and nobody is on track to. He said it hours after a colleague resigned and wrote that the labs are “gambling with our lives.” I have not stopped thinking about it since. Last year I stood on a sidewalk and watched my daughter walk into kindergarten and I cried in the car afterward like a fool. You parents know the feeling. We spend their whole small life building a floor under them, because the world is large and they are not, and holding it back is the one job we actually have. And then the engineers who build this technology tell you there's a one-in-ten chance none of it matters, and the people we elected say nothing and do nothing. We already know how to handle this. We have done it before. We do not let a private company build weapons and sell them on the open market. The Manhattan Project was not private industry acting alone. We do not let a drug company put a pill on a shelf because it feels good about the pill. We make them prove it, in public, with data, under penalty. We do not let an automaker sell a car until someone who doesn't work for them has driven it into a wall. We made those rules because people died first. Every one of them is written in somebody's blood. But a handful of technology companies are building something their own engineers say could end us, and we have asked them for nothing. No safety standard. No independent testing. No liability when it goes wrong. Nothing but the good intentions of an industry that spent twenty years showing us precisely what it is willing to trade for profit. Our children's attention. Our discourse. Our ability to agree on what is true. The reason for the silence isn't complicated. It's money. This industry is spending more to lobby our government than it ever has, and too many people in office have quietly decided that their donors are their constituents. My party has let people down. I'm not going to stand here and pretend otherwise, and I don't think you'd believe me if I did. But look at what the other side has done with total control of the House, the Senate, the White House, and the courts. Their One Big Beautiful Bill let data center developers write off an entire buildout in the first year, while families across North Carolina watch their power bills climb to keep those same buildings cold. Three separate times they have tried to strip states like ours of the authority to regulate this technology at all. First a ten-year ban buried in that bill, pulled out 99 votes to 1 the moment it saw daylight. Then the defense bill. Then an executive order in December and a White House framework in March asking Congress to try again. Understand what that means. They are trying to take the pen out of our hands. Raleigh is the closest thing we have to a government the people can reach, and they want it forbidden from protecting us. Nothing on children and social media. Nothing on ratepayers. Nothing on workers. Nothing on AI safety. That isn't a failure to act. That is a choice, made on purpose, for someone else's benefit at our family's expense.