Why Nobody's Seriously Coordinating to Stop Dangerous AI

An influential voice in the AI world just made a blunt point: people keep saying it's impossible to get countries and companies to coordinate on safe AI, but almost no one actually tries. We treat wars as serious problems worth negotiating. We treat AI coordination like a philosophy club topic — people quickly dismiss it as hopeless and move on. That, he argues, is a huge mistake.

Here's the problem in plain terms: AI is being built in a race. Companies and countries want to win, because the rewards are money and power. But the same race pushes everyone to go fast, and safety can get left behind. If AI systems become powerful enough to cause real harm, an accident could affect millions of lives. Once that happens, it's too late to coordinate.

The author points out something surprising: humans regularly pull off difficult coordination. Boards fire CEOs, then negotiate miraculous comebacks. Rivals reach peace deals after terrible wars. If we can do that, he says, we can probably agree on a basic deal: 'we won't build dangerous AI if you won't either.' The hard part is that AI leaders don't seem to be seriously talking about it, and the public doesn't demand that they do.

So what changes? He wants more detailed, practical conversations about how coordination could actually work — including road maps, hard steps, and alternatives. He wants AI leaders to act like real negotiators with real responsibilities. And he wants the rest of us to stop nodding along when someone says it's hopeless. AI isn't an abstract threat anymore. It's being built right now, and the decisions made in the next few years will shape your job, your privacy, and your safety.