You are absolutely right on that, but you have no reason to assume that they overestimate the problem. In fact, recent changes suggest the opposite. The situation is worse than the scientists have predicted.

This is supposed to be a scientifically oriented site, in which case we should be able to understand actual numbers, not just handwaving.

In a limited energy system, such as an Earth at a fixed temperature, there is a maximum potential value for many of the parameters that we see breaking records. Yes, we would see those records being broken repeatedly, though less and less often as the length of the record increased. In any case it would clearly look like an asymptotic approach towards a limit with each new record on average being closer to the previous one.

Instead we are seeing records being broken on accelerating scales. That's something completely different.

It's a shame you are answering as AC instead of simply setting up a new account. It make it less worth replying to since most times ACs never come back.

I assume you mean this: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2F...

Which only applies to neural networks. However there is evidence, including slime moulds which have no neurons but show intelligence, that intelligence is not just a neural network.

I'd be happy with that, but I think there's a reason that the marketing departments of AI related companies aren't using it.

The important part is the handwaving. In other words, for example

learning or any other feature of intelligence

(my emphasis)

We now have learning clearly separated from the rest of intelligence. Just as we didn't call expert systems "intelligent systems" we should now be calling these "artificial learning" or "learning systems" and not "artificial intelligence" because they are not that.

I don't want the research departments renamed. If they are either searching for how to create artificial intelligence or they are dealing with the outcomes of the research that's fine. What I want is that the products delivered are not allowed to use the branding.

When we call LLMs and related systems "Artificial Intelligence", what we are really doing is false advertising. We need a better name. Maybe "Artificial Skills" and "Artificial knowledge"? This whole AGI thing, pretending that current "AI" is a step on the way to actual artificial intelligence, except in that it's another failed step done by researchers trying to work out what intelligence actually is is a big con job. There's no clarity about that at all.

This is really needed because the systems break in horrible ways, such as Tesla cars being able to drive, but not understanding that they are driving in dangerous conditions where their cameras aren't enough and need to slow down. The confusing this is causing is already ending up with people dead.

They have gone at it far more aggressively than makes sense from a pure cost point of view. There are two possible explanations. 1) they really really care 2) they see this as a strategic technical area where they can get a lead on America and, in the long term either use it for leverage or as a way to give Chinese companies and the Chinese economy in general an advantage over the US economy.

With the US getting more and more trapped into expensive nuclear and coal, it looks more like China is going to get a straregic energy cost benefit.

Unfortunately not entirely true. Ineviably most of us are using a bunch of communication apps like gmail/slack/discord etc. For the ones you use regularly you might want to install the native app, though often these are very close to malware/spyware. For ones you use occasionally or don't trust enough, you want to use the web app and you want to control it's behavior. E.g. you don't want your corporate identity to see your private cookies. In that case Arc does a bunch of useful things whilst dia likely does almost the exact opposite.

Browser plugins don't get access to all the user's data so really don't give you the complete overview this does. It's not close to a substitute.

Oh wait, you were thinking from the user's point of view?? Sorry, nobody cares.

If you committed your workflow to Arc, you are probably wondering how to escape as quick as possible. Unfortunately Arc did a bunch of valuable advances which it isn't totally trivial to replace, but think of it as a lesson in why you should never commit to a product controlled by a company without clear substitutes. Here are some ideas to look into (wiith either open source or at least alternative suppliers) .

* Zen browswer
* sideberry exension for Firefox

What else?

I just bought a pixel device exactly because the custom ROM communities were recommending it. This has always been one of Linux's and Linus's main justifictions for why they were right to stay away from free software puratinism and allow proprietary drivers. Hopefully this leads to a final effort to merge linux mobile back into Linux, to get stock kernels onto mobile devices and to remove Google from the ecosystem. We need a new bitkeeper moment.

We obviously can regulate AI to some extent and in some ways. A very simple, useful, thing would be to simply know when results have been generated with AI and to allow people to opt out. Yes, people will cheat, but provided you have strong regulation where they go to prison when that causes serious consequences, they will also get caught occasionally through accidents or clever investigation and that will provide a clear deterrent. Providing whistleblower protection and bounties will make this very likely to come out.

I wouldn't count on this. I have noticed that a lots of web sites connect to localhost. I always assumed it had some semi-malicious reason. Even now I notice that slashdot is connecting to 1248321188.rsc.cdn77.org. I'd assume that this is basically the same idea.

In went $1.5 billion. Out came $37 million they probably had a payroll of, say $20 million. If someone managed to keep the rest and gets away with it they may well consider themselves pretty clever.