A lot of people are extremely concerned about the prospect of AI superintelligence, and the possibility that it could supersede and perhaps even destroy humanity.
I spent a lot of the past four years or so researching this exact topic.
But increasingly it’s coming to look like AI is actually shit. In other words, it’s not going to take over the world and send Terminators to kill us (what I call the Skynet Complex). And nor are we going to luxuriate in Star Trek-style indolence while AI does all the heavy lifting.
AI can replace *some* human performance, but as we are learning, it’s usually inaccurate, unreliable and expensive. Most companies who’ve used AI have lost money on it.
But as Cory Doctorow notes, AI is an enormous danger. It’s going to destroy our global economy as it eats up all the investment and provides almost nothing except environmental degradation and water shortages in return.
It’s a combination of the world’s biggest nothingburger, tulipmania for the digital age, and the displacement of people in work by unreliable, fantasising, digital plagiarists.
We should stop AI now, not because it might kill us like Terminators, nor because its creators stole all their training material, nor because it’s an environmental disaster. But because we’re betting the global economy on a horse with three legs.
