At some point, someone is going to die because they followed an AI’s advice, the companies involved are going to get sued and laws will be passed making companies criminally liable for this kind of abject misinformation, at which point this entire sub-industry will evaporate. It would be nice if that happened before somebody dies, but probably not.
The precedent has already been set in the Air Canada AI case - companies will be held legally responsible for AIs that lie to their customers. Any smart company should get out of this bullshit before they’re dragged through the courts, not after.
