I disagree with a lot that this essay says, but I just want to hone in on one practical thing:
> Noticing that AI companies put the world at risk is not enough to support greater government involvement in the technology.
If government (which, let's not forget, is us) doesn't step in to protect the populace from abuse, then who will? We certainly can't trust the AI companies to do this. I don't accept the premise that we just have to remain defenseless.
We certainly can't trust the government to either, depending on which government it is. Will the government - generally the most dangerous entity in a country, as it has the military - start by regulating itself?
Voting works in a sense that when society is in a "somebody ought to do something about it!" panic mode over some issue, it can make "something" happen.
It is very rare for that to actually translate to good policy, though. If you're lucky, best case, you get a crutch that treats some of the symptoms. Worst case, you get a knee-jerk reaction that makes everything worse.
> The most dangerous entity in a democratic country is a widespread belief that voting doesn't work.
I don't know that that's true. Voting is so low resolution a signal, between two options, on 10000 issues, that it is pretty meaningless in terms of policy. You could say it works in the sense that your team gets in, though.
Perhaps, but equally populations are growing, so each person's signal gets less and less valuable over time. Democracies birthed in places with a high sense of buy-in and relatively few people voting would've felt very different to how things feel today, and I don't know how many ways there are to rectify that.
I work for the government, not federal but local. While I certainly don’t trust any level of the government I certainly trust them more than I trust Microsoft, Open AI, or any other SV billionaires or wanna be billionaires.
Yeah, I understand where this author comes from but at the same time the whole premise sort of garners a head-tilt, half-squint, visible confusion. Companies only really self-regulate by way of failure. If they die they prove the negative hypothesis, but the positive hypothesis was just "this product will make net profit" not "This product will not materially harm society".
Among the things that the article omits is the question of upholding the existing canon of laws in the face of AI.
As far as I see, much of European AI regulation tries to contain fallout where existing laws are under threat to be undermined (i.e. authorship, safeguarding public opinion, competition, harm/safety etc.)
16 comments
[ 3.4 ms ] story [ 47.6 ms ] thread> Noticing that AI companies put the world at risk is not enough to support greater government involvement in the technology.
If government (which, let's not forget, is us) doesn't step in to protect the populace from abuse, then who will? We certainly can't trust the AI companies to do this. I don't accept the premise that we just have to remain defenseless.
It is very rare for that to actually translate to good policy, though. If you're lucky, best case, you get a crutch that treats some of the symptoms. Worst case, you get a knee-jerk reaction that makes everything worse.
I don't know that that's true. Voting is so low resolution a signal, between two options, on 10000 issues, that it is pretty meaningless in terms of policy. You could say it works in the sense that your team gets in, though.
As far as I see, much of European AI regulation tries to contain fallout where existing laws are under threat to be undermined (i.e. authorship, safeguarding public opinion, competition, harm/safety etc.)
Do better.