James Barrat doesn't seem to know what he's fighting against. He thinks alarming people about the dangers of technology would halt the use of AI. But he's wrong.
Considering that corporations do literally anything to make more profit, no matter what people think about them, the ones that should be persuaded are them.
Take Google for example. Google uses deep learning(I'm not sure if this counts as AI) to improve their once-crappy (and still-crappy) translation service, to show "personalized" ads, and keep people watching YouTube. For that, Google peeps into our computers via Chrome. Chrome is notorious for it's terrible RAM usage, but also for how it stalks people. If Google only uses them for training their AI's, it may not be a big problem. However, we don't know if Google is doing just that or also sending information to NSA. Since Mozilla Firefox has gained much attention during these recent years, it seems like people are very concerned about Google.
It's easy to persuade people that AI is bad. You just throw them a plausible cliché-filled scenario where AI tries to destroy humanity (example: i-Robot(the movie, not the novel)) and that's going to make the headlines. But as long as AI means more $$, companies are not going to stop.