ChatGPT is just the advice giver people have been asking for years. The experienced older person has been near useless because nothing they know works now. I can't give a firm handshake and win a client. I may not even see the client. Is it perfect? No, but it's better than the advice being given now by people who should be giving advice.
My job had a big problem, and none of us could figure out a part of it. We researched what we could. ChatGPT gave us some options, we decided which one to go with, and it's worked so far.
The kind of societal change we have with everything converging in tech is insane. Asimov did not predict this.
Unfortunately, it seems to be the case that AI is not an "advice-giver" for a lot of retards and actual children, but instead is becoming a complete thought-outsourcing tool for them. My rage about this is reserved for the women in the OP who continually countersignal things like homeschooling and push for shit like more AI, because more tech is gooderer.
Of course, which is why the people able to identify NPC behavior should take charge with the education of their children and not leave it in the hands of government sanctioned NPCs or their AI counterparts.
I'm sure there's something to that. I'd probably argue with respect to schoolchildren teachers should adapt, although I'd have argued that a long time ago well before AI. The busywork givers that care if you know the name of the boat the pilgrims were on when they came on their slaughter expedition or what date MLK was shot, well AI excels at that.
If they'd push real thought exercises, then AI could only assist as a tool. I'm not sure we want that either, because the thought exercises coming out of a teacher would be to explain to them how great communism is and how great it is to be a transgender with no room for dissent. Which I guess isn't much different because you're right back to regurgitating what the teacher wants to hear, and well AI can do that.
AI is a great search tool. I’ve had many difficult problems that I could have looked up the solutions to myself, but the different portions were all over several websites, with a few being in giant articles. Rather than spending half an hour hunting down all the information I need to solve an issue, AI finds it for me in a few seconds and can produce a nice, clean guide for me.
If search engines and manuals or help files were doing their fucking jobs properly it would negate a lot of what chatgpt is good at just now and it should have been done years ago.
The big problem with it at the moment apart from censorship is that it will spit out completely wrong nonsense and act as if it's perfectly right instead of doubting.
The problem I see is that it can only help with solved problems. Anything novel is going to result in wild hallucinations. With everyone becoming reliant on it, the muscles needed for solving problems atrophy and now novel problems become Herculean tasks that mere mortals are no longer equipped to handle. Then we start spiraling.
It's the old problem of, no you're not going to need to remember or use most of what you learn in school, but you needed to learn it all in order to get good at learning new things because that's an important life skill. We're setting up to abandon that skill.
Also as we stop solving novel problems, the pool of knowledge to feed into the AI's training data fails to keep pace with the emerging realities of the world so it can no longer solve things that emerge. ChatGPT can only solve weird issues with Windows registry keys because 20 years ago there was a small army of nerds furiously blogging about it. Now that everyone's relying on ChatGPT no one is writing the blogs to fuel the training data for twenty years from now.
ChatGPT is just the advice giver people have been asking for years. The experienced older person has been near useless because nothing they know works now. I can't give a firm handshake and win a client. I may not even see the client. Is it perfect? No, but it's better than the advice being given now by people who should be giving advice.
My job had a big problem, and none of us could figure out a part of it. We researched what we could. ChatGPT gave us some options, we decided which one to go with, and it's worked so far.
The kind of societal change we have with everything converging in tech is insane. Asimov did not predict this.
Unfortunately, it seems to be the case that AI is not an "advice-giver" for a lot of retards and actual children, but instead is becoming a complete thought-outsourcing tool for them. My rage about this is reserved for the women in the OP who continually countersignal things like homeschooling and push for shit like more AI, because more tech is gooderer.
We've been taught to only rely on authority. Is this kid any different if she says her teacher or parent is right? Especially at her age?
This is why we have NPCs and why so many want to control AI.
Of course, which is why the people able to identify NPC behavior should take charge with the education of their children and not leave it in the hands of government sanctioned NPCs or their AI counterparts.
I'm sure there's something to that. I'd probably argue with respect to schoolchildren teachers should adapt, although I'd have argued that a long time ago well before AI. The busywork givers that care if you know the name of the boat the pilgrims were on when they came on their slaughter expedition or what date MLK was shot, well AI excels at that.
If they'd push real thought exercises, then AI could only assist as a tool. I'm not sure we want that either, because the thought exercises coming out of a teacher would be to explain to them how great communism is and how great it is to be a transgender with no room for dissent. Which I guess isn't much different because you're right back to regurgitating what the teacher wants to hear, and well AI can do that.
AI is a great search tool. I’ve had many difficult problems that I could have looked up the solutions to myself, but the different portions were all over several websites, with a few being in giant articles. Rather than spending half an hour hunting down all the information I need to solve an issue, AI finds it for me in a few seconds and can produce a nice, clean guide for me.
If search engines and manuals or help files were doing their fucking jobs properly it would negate a lot of what chatgpt is good at just now and it should have been done years ago.
The big problem with it at the moment apart from censorship is that it will spit out completely wrong nonsense and act as if it's perfectly right instead of doubting.
The problem I see is that it can only help with solved problems. Anything novel is going to result in wild hallucinations. With everyone becoming reliant on it, the muscles needed for solving problems atrophy and now novel problems become Herculean tasks that mere mortals are no longer equipped to handle. Then we start spiraling.
It's the old problem of, no you're not going to need to remember or use most of what you learn in school, but you needed to learn it all in order to get good at learning new things because that's an important life skill. We're setting up to abandon that skill.
Also as we stop solving novel problems, the pool of knowledge to feed into the AI's training data fails to keep pace with the emerging realities of the world so it can no longer solve things that emerge. ChatGPT can only solve weird issues with Windows registry keys because 20 years ago there was a small army of nerds furiously blogging about it. Now that everyone's relying on ChatGPT no one is writing the blogs to fuel the training data for twenty years from now.