I laugh a bit at this because I like Marvel movies and his AI was fun. However, as you said, it's still just fiction. Because it works that way in a movie doesn't mean it will ever work that way in real life. It might end up being similar, or it might not. It just depends how things with AI progress over the next years. And you say it doesn't make him stupid, but again, it is a movie and they can make it be however they want it to be. Another movie could make it so the AI use did make him stupid. That's entirely up to the writers and isn't necessarily a reflection of reality.
What we do see in real life are students who are using AI to do their work for them. This does often lead them to not know how to do the work themselves or to not know the material they are writing about. That isn't going to be true for everyone, of course. Some will use it to assist them instead of doing the work for them, and they are likely to still know what they are doing. It is also true that before AI, you could run into this same situation of not knowing how to do the work or not knowing the material because of using Cliffs Notes or paying someone to do it for you. So the issue isn't entirely connected to the use of AI. But the use of AI does increase the number of people who are relying on it to do everything for them, which does definitely lead to more people not knowing how to do things themselves. It is easy for someone to just go to AI and say they need a paper about X written and have it written in a few minutes. But if they want to understand X, that takes much longer and if they aren't willing to write the paper, they also are less likely to spend time learning the material either.
As I said before, I support using AI to assist with things. And I think there will be a future where it can be used to do significant amounts of work and still be accepted by most people. But I we aren't there now. As you've seen here, even those who aren't against AI are still hesitant about its use on any larger scale. It just isn't ready for that yet. It will be. And more people will start to accept it eventually. But that isn't where we are today. As I said, in 5-10 years, things may be different. For now, it can be a useful tool to assist with small things, but it isn't something that will do all the things you suggested in the OP. As was shown earlier, even that AI says it's not going to be used for those things. And also as I said before, TFP already uses AI for at least some minor things, and perhaps for more than we know. So any benefit that AI provides is, at least in the areas they are using it, already being seen.
Anyhow, I think if you want to have an AI conversation, it's fine. But I think it should be about all AIs and not focused on a single AI. It is against forum rules to advertise stuff. You said you were paid, though it's hard to tell if you're joking or not. If you were paid, then you are breaking the rules with this entire thread. And even if you weren't paid, it would be better to discuss all AI anyhow. Anthropic AI (the correct spelling of it) may be better in some things than other AI, but worse at other things. AIs can be designed with a specific focus, based in part on what material they have the AI "learn". So some are better at some things and others are better at other things. Not focusing on only one AI lets you have a more diverse conversation about AI use rather than focusing on a single AI that is new and is less known than others.
I would also suggest not being so unwilling to admit to the shortcomings of AI use. It is a fact that AIs have many limitations to what they can do. Acting like there aren't any limitations and that AI is some miracle cure makes your arguments weak. It is different if you instead agree when someone points out an obvious limitation and then give a reason why you think that may not be as much of a limitation or that it is something you feel will be improved soon or whatever argument you want to make. Acting like you know everything about the subject and telling others they know nothing if they disagree with you and then completely ignoring factual limitations of AI use just makes what you say not really worth reading. I even asked you if you have personally programmed with AI and debugged AI code and such and you didn't respond. That suggests all your talk about it is just what you've heard or read and not from any practical knowledge of using AI in programming. Consider that a number of the people here who have responded actually have used or tried to use AI in programming in one form or another, either professionally or personally. If those of us who have done so say that it's not time to start relying on AI for programming, maybe you should listen.
.
. It doesn't make him stupid, and it doesn't steal or destroy human work and discoveries. Do you see the analogy here? (It's fictional, but isn't this fictional AI, an artistic creation, what we would most like our AIs to be as close to?)