Bernie Sanders floats banning super intelligent AI.

 

Senator Bernie Sanders. Politico/Getty.

Senator Bernie Sanders has floated legislation that would ban super intelligent AI. Politico. The ban would punish anyone who developed a super intelligent AI with 20 years in prison. Critics say such a ban would handicap the United States against China. Sanders has also called for a moratorium on data center construction. Such legislation would face strong headwinds in Congress and might be ruled illegal by the courts. However, the ban and moratorium might be popular with voters deeply skeptical of AI. 

My Comment:

There's very little chance of this passing. Not only would the tech companies not go along with it, the executive branch would have a fit as well. There is zero chance that the United States would simply wave the white flag to China and allow them to develop advance AI while leaving ourselves without similar technology. It would essentially be suicide. I think there is less than a 1% chance of this legislation happening and even if it did, I think the courts and the executive branch would prevent it form going into effect. 

Not that I think a true super intelligence, in the way sci fi people define it, is likely any time soon. AI can, and already does, some things better than humans, but they are not sentient and still make major mistakes. Such a thing might be possible in the far future, but I am skeptical in the short term at the very least. I generally think it's another example of the Whig view of history, where people assume that society will keep advancing and there will never be any regression. I think we absolutely could end up with hurdles that will take decades or even centuries to overcome when it comes to AI super intelligence, if they ever get solved at all

Do I think there are dangers to advanced AI? Sure, there are dangers with every technology. Many of them are not predictable before the technology is released. But the AI safety "sci-fi tech uprising" that ends humanity? That's not one of them. AI could do a lot of things, many of them bad, but I don't see them gaining actual sapience, which would be required for malice. 

Regardless, Sanders is a savvy political operator and I don't think this is about super intelligent AI. Sanders is a communist (ok, a socialist), and his objection isn't that AI may go rogue and kill us all, it's that companies are making a profit. He's just exploiting popular sentiment that is critical of AI. 

So why are people critical of AI? I don't think it has a thing to do with AI safety. Much of it is simply fear of change. There are always winners and losers when a new technology is rolled out. Horses lost out when the automobile was invented, industrial workers lost out when robotic automation became common and now a lot of entry level white collar jobs are going to disappear as well, and many already have. It's pretty natural to be opposed to a chance when it does in fact harm people, but that also usually doesn't do a thing to stop progress. 

The other problem is that the current tech barons don't understand the concept of noblesse oblige. The old robber barons of history were also centers of progress and made their millions by changing the world. But unlike the tech barons, they understood that in order to keep the loyalty of the people, they had to give something back. So they built museums, parks, colleges and charities. 

Today's tech barons? They don't do any of that, at least not for Americans. Many of them send their billions to the third world, not really caring if they have a positive effect or not (a good example is the mosquito nets they buy being used for overfishing). Even Elon Musk, who is probably the most benevolent of the tech barons, hardly gives anything back to the people, most of his spending is political. To be fair, he does give folks access to Starlink during disasters, but that's the exception that proves the rule. 

In that, I guess I have to grudgingly admit that Sanders might have a bit of a point. If we want AI (and I think we should), I do think that these tech barons have to shape up. Sander's solution is idiotic of course, but I would argue that the tech barons should indeed be giving back to the people out of naked self interest if nothing else. I don't think people would be anywhere near as opposed to AI and data centers if the tech companies cut a check to improve their lives.