After all, asserting the technology’s world-altering danger is also a way of emphasising its world-altering power, which, in turn, feeds the industry’s insatiable need for further investment. Saying AI “could kill everyone” is, in effect, a way of saying it’s more consequential than any other thing that humans have ever built (with the possible exception of nuclear weapons).
But the real utility of the apocalyptic narrative for the industry is that it distracts attention from more tangible issues over which democracies could have leverage. Such as what? First, there’s the monopolistic power of tech companies, something for which existing legal tools are readily available, provided governments have the political will to use them. Second – even more straightforward – we could just make AI companies legally liable for any harms that their products inflict on us humans. And then we could leave apocalypse to the movie industry, where it belongs.
@ChrisMayLA6
There is a very old rule of law that someone who has created something potentially dangerous on their land that was not naturally there, is subject to strict liability if it escapes & harms a neighbour's property or person, without need to prove any fault. (I believe the analogous law in England is "non-natural use of land").
So why not treat this particular novum opus under the same ancient principle?