However, your subject line is tactless....

With the difference that usually the use of open source code cannot be checked or monitored, and OpenAI can easily monitor all conversations with their chatbot. Don't tell me it's too much effort, they're an AI company. And if they really can't, then we all know what their technology is really worth...

I know bashing on parents is in vogue, but this type of law is necessary for different reasons.

Of course the ignorant (you know who you are) will say this is the parents' business. First, helicoptering isn't a solution either, and no, there's no serious control anyway.

My main point however, is that this kind of law makes it impossible for schools and such to prescribe WhatsApp and such as a communication channel. So when our eldest kid got into middle school, after the government had decided that WhatsApp was only for 16 and up, before 16 no go without parental consent, no discussion was necessary.

So it's not about parents not parenting, it's about enabling parents to keep others from prescribing stuff that would leave their kids behind if they would decide to make that stand.

You're absolutely right, but it's not even necessary to look as far as experience. Just start out with senses, and a drive to survive. Or let's say, a sensation of discomfort before any risk to survival is even on the horizon. Humans normally try to avoid discomfort and will look for modifications in such a situation.

In the 90s in university I already joked: give me a computer that can feel pain, and I'll give you artificial intelligence.

I think you could have broken all your phones with a single Nokia 3210 which would have still worked fine after the abuse and would have to be replaced solely because the 2G networks got deprecated...

Perhaps an AI spelling checker would have caught that error........

Which ones are currently up to par, in your view? I'll try out my standard litmus tests... And yes, I'll happily take back my words.

Sure, Chat-GPT is still very young, but according to credible sources, LLMs have practically reached their limits, there's not enough data to train them more. And considering the kind of mistakes they make, even more data isn't going to get them to truly understand anything. So extrapolating from there isn't logical. Now if there's something realty new that can combine with LLMs into something really next level, maybe. But I've not seen any such thing yet.

Hah! I hadn't thought of that one. The first time I saw the agile manifesto, I thought it was in jest. But then, I'm a hardware designer, did IC design at that time.... Still not entirely convinced it isn't.

Good point. So only if you turn that off (the developers who have access to that option), will an LLM respond with the exact same answer to the same input.

I'm not saying AI isn't going to be disruptive, but I've had a long look at LLMs and they're not much more than clippy (or autocorrect) on steroids.

So there's an increase to be had with respect to the efficiency of many types of desk work and then some. Since proofreading and checking stuff will still be required, although of a different nature (not spelling errors but entire missteps and such), there's not a 100% reduction of actual desk work, perhaps 50%? But anyone doing desk work knows that producing text or code is merely a part of their job. You need to figure out what's going on and what that means for your upcoming work. LLMs are a great parlour trick but they aren't going to help much with that. Or at least, I've not seen it capable of doing that. So, big whoop, people will be more productive.

As for other AI, and their impact, sure, being it on and we'll see. For now I'm tired of the hype. Allow me to repeat myself...

On the hype about AI being oh so capable, and going to put everyone out of a job:
When the first AI company CEO lets their AI handle all their financials, I'll start paying attention to the hype. But not until the AI programmers let their AI handle all their financials will I start believing. Indeed, convince me by putting your money where your mouth is. Until then, i won't pay a lot of attention to these AI stories.

On general AI / artificial general intelligence:
I'll believe someone managed to develop full general AI when the news hits of the first company with zero employees, just shareholders / owners.

Thanks for letting me get that off my chest.

But you are still giving an AI system that can "hallucinate and produce unexpected outputs" (Microsoft's words, not mine) full access to your personal files.

Nope, LLMs don't hallucinate. Their algorithmic output is deterministic and just gives the output based on the input, the training data, and various settings and configuration values.

So (repeating myself): AI don't hallucinate from time to time. Every answer they ever give is equally made up. What people call hallucinations are merely cases where the made up answers are ostensibly wrong.
Any AI may apologise when it's pointed out that their answer was incorrect, even if in fact, it happened to be correct.

Happy to be of service, have a nice Sunday!

Or make it hit Belarussia, and hilarity ensues. Or, make it hit Poland and get NATO involved. No more jokes, but at least talks about Ukraine losing territories to Russia would be off the table...

I scanned the article and all current comments, and I can't find a single instanciation of the word AI...