So move them to offices in Thailand or Viet Nam or a bunch of other places. You do not have to contract out the work, just create a local subsidiary.

So, instead of hiring these foreign workers and bringing them to the US, they will hire them and employ them in their home countries at even lower wages.

I see that there are people questioning whether he was deserving. I have no opinion on this, because I have no idea who else was nominated. Before criticizing a particular award, you should go back and look at some of the previous recipients. The Nobel Prize committees are just as fallible as any other committee system.

I was replying to the previous post, which refers to an ax murder. Exactly what laws have they broken? To break a law requires both an act and intent. Since AIs are not legally people, they cannot have intent. You would have to argue that their creators either had intent to break the law (very doubtful) and were willfully negligent (very difficult to prove).

I chose that point because that was the example that the person I was replying to chose. The basic point remains to prosecute someone you have to show that they have violated the law and also either intent or negligence. Proving negligence can be difficult and the AI companies can afford good lawyers. So far, prosecutors have not thought it worthwhile to act.

As far as we know, none of these programs have killed anyone yet. If they actually break a law, the company using them should be held responsible. Unfortunately, most of what it is doing is not illegal although it may be immoral.

Which is why string theory has not led anywhere useful.

Yes, virtually everything has an upside and downside. The question becomes, is the upside greater than the downside? So far, I am failing to see much upside from AI other than making a few people richer. The downsides are all around us. I do not think that problem is the concept of AI, so much as the way it has been implemented. I would support deleting all the existing models and starting over with a better development plan than feeding your model all the junk you can find and hope that it can locate the diamonds in the shit pile.

Who will score them, other police? The basic issue is that police, prosecutors, etc. are judged on how many arrests or convictions they get, not on how many are justified. It is more important to them to make an arrest in a case then to actually identify the guilty party.

We sincerely hope.

I believe that you are correct. If the infection occurred because of a containment failure in the lab, the pneumonic form would be possible. Neither is good, but the pneumonic form is far more contagious, since it can be spread through the air.

Plague is endemic in rodents in a number of places. It is probably native to the Russian parts of eastern Asia, but it is present in ground squirrels in the western US as well. Fortunately, it is usually easily curable with standard antibiotics when humans contract it, which they do. In Russian Asia, which is sparsely populated and poor, it could be a significant problem. An antibiotic-resistant plague, which given the tendency of bacteria to develop resistance is not impossible, would potentially be a disaster of major proportions.

True enough, but what else can they do? If you start assuming that the basic "constants" of the universe are actually variable, you end up with so many adjustable terms that your model could predict anything.

That is $2 trillion, and that is just the US. I think even META would have a hard time coming up with that kind of money. Of course, I doubt anyone would miss them, if they went out of business, but still, it seems silly.

Agreed. An AI developed specifically to play a particular obscure game (even if a very complicated one) seems both highly limited and dubiously useful.