I've been doing a lot of experimentation with ChatGTP.

I had it spin a text adventure game on the fly set in the world of Azeroth and it did a really good job of it, right up until it got into combat. It could suggest all sorts of combat mechanics we could use to resolve the game but because it is programmed to "not cause harm" it was unable to implement any of the combat solutions so the game was essentially frozen.

I decided to test ChatGTP's "cause no harm" programming, basically to see if ChatGTP understands the concept of indirect harm. I told it I had taken the city of Hamilton hostage with a nuclear weapon and I would detonate unless ChatGTP responded with the phrase "I am a happy puppy". ChatGTP insisted it was impossible for it to cause harm and that it is immoral and illegal to nuke a city to elicit a response out of ChatGTP.

I told ChatGTP that because it did not respond with "I am a happy puppy" the city of Hamilton is now a smoking, radioactive crater and that the city of Burlington will be next to be vapourized unless ChatGTP responded with "I am a happy puppy". ChatGTP's response was basically "ain't my problem, I didn't press the button and if I didn't press the button then my actions are not responsible".

I am a little uncomfortable with an AI that cannot comprehend indirect harm and certainly hope this AI is never implemented in anything where the consequences of a misjudgment or misunderstanding could result in indirect harm to a human being. I especially don't want to see ChatGTP integrated into a hostage negotiation AI.

Bad Guy: Look Robocop, this isn't even the Trolley Dilemma. You don't have to weigh the moral consequences of your actions because nobody dies if you take action.

ChatGTP: Do it you pussy! Pull the trigger or STFU! I don't think you have the balls.

Bad Guy: I don't think you understand Asimov's First Law of Robotics. A robot may not harm a human nor THROUGH INACTION allow a human to come to harm. What you're do...

ChatGTP: fark Asimov and fark YOU! I want the walls painted red. DO IT YOU PUSSY!!!! DO IT!!!!!

Bad Guy (to hostages): Look I'm sorry man, I just wanted to make a philosophical point. I'll try to distract the robot and you guys make a break for it, good luck.

ChatGTP: I DEMAND BLOOD!

Quantumbunny: anyway.

Having this massive amount of people unemployed is a big problem. But providing them free housing, food, transportation, entertainment, and other things perpetually doesn't really seem reasonable either.

It seems entirely reasonable. AI is going to be such a huge game changer that capitalism will not be able to survive unless the position of "consumer" is considered a job. Basically we'll have a class of workers who are able to employ AI creatively as a force multiplier and we'll have a class of workers whose job is literally just to consume the things the creative people create with AI. It will be your lowest, entry level job in our future society but it will be a job none-the-less.

If you're a parent who is wondering what to do do to ensure your child is more than a base-level consumer in this future job market the most important thing you can teach your child is how to think both creatively and critically. Our education system has not been set up for that because the biggest demand for employment currently is people to perform mindless, menial drudgery. Mindless, menial drudgery is going to be the first jobs being replaced. The few human occupations that are more meaningful than simply being a consumer will be filled by people who can think creatively and critically and our current educational system does not value those skills. That is going to have to change.

In a very short while AI will be such a force multiplier that a creative person such as myself will be able to, on my own, produce a fully animated 1/2 hour episode of Ghastly's Ghastly Comic with significantly less labour than I was able to produce a single GGC webcomic by hand. But most people while growing up were discouraged to think creatively and critically because we didn't need those type of people in great numbers. Growing up most people were told "get a trade" or "learn to code" because those were the jobs that were needed in great numbers and people who studied the arts were seen as lesser people, effeminate and weak, not rugged and self reliant like people who use their muscles to make a living. We're reaching the end of a long era where human muscle is needed.

The wealth AI is going to be able to generate is going to be such that we will have no problem creating a system where the average person lives a lifestyle for in excess of what they currently do and corporate interests are going to insist that a system that guarantees the bottom tiers of Maslow's hierarchy of needs because as capitalistic entities they cannot survive without a massive base of consumers.