Reminds me of the scene in Charlie and the Chocolate Factory.

Programmer: Gentlemen, I know how anxious you've all been during these last few days. But now I think I can safely say that your time and money have been well-spent. We're about to witness the greatest miracle of the machine age. Based on the revolutionary Computonian Law of Probability, this machine will tell us the precise location of the three remaining Golden Tickets.

[He pushes buttons on the machine, which prints out a response]

Programmer: It says, "I won't tell. That would be cheating."

[He pushes more buttons on the machine again]

Programmer: I am now telling the computer that if it will tell me the correct answer, I will gladly share with it the grand prize.

[the machine prints out another response]

Programmer: He says: "What would a computer do with a lifetime supply of chocolate?" [getting increasingly agitated, he begins mashing the buttons] I am now telling the computer EXACTLY what he can do with a lifetime supply of chocolate.

> Internet via satellites is cool, but still can't compare itself to the one provided by optic fibers, so not a game-changer.

As someone whose family lives in a rural area, it absolutely is. The difference between Starlink internet and the crap they used to get from Viasat and other satellite internet companies is like going from dial-up to fiberoptic.

Not to mention use cases like natural disasters (it was used effectively in areas suffering forest fires, where internet lines were down or didn't reach those areas), underdeveloped countries, underdeveloped areas in developed countries, etc.

Heck, being able to use Starlink while traversing the ocean would be a game-changer in and of itself for freight, military, and passenger ships.

Yeah, the 'post-thinking machine' society in Dune wasn't any better.

> Predictive policing... wouldn't that require preemptive due process?

Depends on how you do it.

For example, the model might say "a crime is likely to occur on this street on these days". So, assign a police car to hang out there on those days - you use the presence to prevent the crime from occurring.

Asking it for factual information is extremely hit or miss.

My primary use for it has been as a brainstorming assistant. Bouncing ideas off it, asking for variations or alternatives, that sort of thing.

That's still pretty amazing. I'm definitely trying this out - I don't DM DND, but it will probably be helpful for brainstorming when writing a paper or a novel.

Did the AI just reinvent the chocobo?

As a child, my primary use of the internet no filter was a mix between Pokemon fanfiction and looking up adult websites when my parents weren't home. So probably that.

And most likely many many stupid questions.

Which is incidentally what I frequently use it for now as an adult. I spent an entire evening asking ChatGPT about the pros and cons of owning rabbits as a pet. Then I had it put together a DND campaign. Then I had it give me some examples of emails I might send. That last one was really useful, to be honest.

We don't even have a hard test to determine people's age for adult websites, but you want one for a chatbot?

Oh, fair point. I thought you meant about communicating with AI in general.

I mean, [Her](https://en.wikipedia.org/wiki/Her_(film)) went pretty okay.

I hope you're right. I've never seen anything good happen when people start screaming 'think of the children' about new technology. I'll check back in with this thread in a year, see how things have gone.

>My concern is around children. A disclaimer would not help.

Then I'm still questioning what you think would help. Your suggestions so far have been to imbue a computer program with professional judgement, an understanding of morality and ethics, and safeguarding training.

If you know how to do this, you've already invented AGI.

>I don't think we need 15 years. Maybe even 1 is enough. What I am saying is when it comes to children a lot more safety work needs to happen.

You're more optimistic than I am. My expectation is that there will be a largely symbolic uproar because some kid was able to Google "how do I keep a secret" by using a chat program and nothing of any actual benefit to any children will occur.

It's two tweets down in the same thread by the same guy. Did you finish reading what you linked?

In my experience, ChatGPT very blatantly presents itself as a computer program. I've asked it to invent a fictional race for DND and it prefaced the answer by reminding me it was a computer program and has no actual experience with orcs.

If your concerns would be met by the program beginning each conversation with a disclaimer of "I am a computer program and not a real life adult human being", then I'm perfectly fine with that and support your idea.

If your concern is that a chat program needs to be advanced enough to have "moral and legal" judgement, well, I guess you can come back in 15 years and see if we're there yet.

All advice is unchecked. Learning to be critical of advice is a wonderful life lesson for children to learn.

The fact that you're calling it an 'AI' instead of 'sophisticated chat program' is the real issue here, honestly.

Not really, no.

I'm in my late thirties. I have no idea how old you or anyone else on Reddit is. You have given me no background check or safeguarding training. Some people in this thread might be kids, I have no idea.

Kids use each other as confidants. Do you background check the other 12-year olds?

Kids know how to use Google. What is the fundamental difference between going "How do I hide a bruise?" to a chat program and searching it on Google?

I think this is a knee-jerk reaction to an interesting new gadget and that there is literally no solution to the problem you are perceiving.

Consider the issue shown in the Twitter you linked. How would you fix this? Cause the chat program to shut down if you admit your age is under 18? Prevent it from responding to questions about bruises or physical injuries? Give the program a background check?

It rapidly disperses in the sense that the gas scatters too much to actually influence anything's orbit. For you to use it to clear out any swath of space, you would need an enormous amount of gas released at a high velocity. And even then, since all the debris is constantly moving, it's not like that area is now 'clear'. You just have slightly less debris at that orbital inclination. Or slightly more debris, depending on the design of your gas bomb.

I mean, that is worrisome, but not for the reason you're implying.

This is how technology gets neutered to the point of complete uselessness.

"A program that can answer questions? But what if a child asks questions! They could ask any question at all and be given answers, even if the contextual nature of the question makes it inappropriate in ways a program can't possibly understand! Quick, it must be destroyed! Destroyed immediately for the sake of the children!"

I'm reminded of how people were worried that kids playing Dungeons & Dragons would result in them sacrificing their friends to Satan. What the heck is stopping the kid from googling "how to hide a bruise"? Literally nothing. I just did it, the first result is a 'how to' video on YouTube so I can be shown how to do it properly. Yet somehow this chat program is a horrible, terrible menace.

> You cannot trust any information or advice it gives you, hell you can convince it that 1+1=3

So, if you give it incorrect information, it provides you with incorrect information?

I am shocked, shocked, to be told that a computerized system operates on the principle of "Garbage in, garbage out".

> The article mentioned it, but didn’t offer any good solutions to the “menial, but necessary” jobs problem as they put it.

Automation?

If my basic needs and wants were met, I would absolutely stop working. I'm hoping to retire in about 6-7 years and then never work again. Some people in my same situation are already talking about how they'll get another job to 'keep busy'.

I'll keep busy, but I'm not going to work. I can be plenty busy by reading, playing video games, watching Netflix, spending time with pets, and sleeping in on weekdays.

That's exactly what I would do with a guaranteed income that covers my basic needs. Now, $700 a month would not be that. That honestly would not really change my life at all, though I know it would for many people.

But it wouldn't erase my need to work in order to support myself financially, even though I have no inherent interest nor desire to work.

I dunno, the Roboroc probably cleans better than kids.

I program mine to run while I'm at work. Sometimes I'll come back and it'll be stranded somewhere, but 90% of the time it's back recharging in it's cradle.

There are so many potentially interesting applications outside of "make a digital ghost of your loved one when they die", which is (in my opinion) the most boring possible use of this technology.

You could digitize a copy of noted scientists or experts in their field and replicate them in schools as teaching aids. You could create a digital duplicate of yourself and talk to it. Trying to make a tough life decision? Generate five digital copies of yourself and have a discussion on what to do. Or boot up a copy of a noted philosopher or therapist. Be a game designer, hire people to upload aspects of their digital avatars, use them as the basis for your NPCs.

Ah, that's true. Probably not, then. But one day! One day.