I appreciate this perspective. It's quite interesting to see what someone who is a generation above me thinks about all this. My personal stance is that I believe the chance we can build AI which is truly beneficial to humanity is not that bad. While it's extremely hard to develop processes where its provable that some sorts of specific human values will get instilled into an AI, it doesn't seem nearly as hard to do it in practice and so I think there are approaches which could work out well enough, even if its not provable that an AI is going to behave under all circumstances beforehand. What makes me think this is how seemingly well our limited approaches thus far have worked to get AI models which generally do what people want them to, leaving me to think that it is a viable strategy to make an AI which would be more capable than humans at figuring out how to build more aligned models which would try to accomplish that task and not explode in humanities face as long as nobody is using it to specifically cause harm. Something like this could give us models which cause relatively minimal problems even when people try to use them for harmful purposes. Well, at least I hope a strategy like this can work out, but it seems to be roughly what AI labs are doing and the current approaches have thus far worked better than I would've expected them to. Of course it could go wrong in all sorts of ways, but if the researchers are any more capable and reasonable than me, I feel like things should work out decently well.
What everybody call AI today isn't really AI. No original thought, no creativity, no opinions, no drive. It's just another human driven tool. The fact that many will lose their job is nothing new.
It sounds like you've watched one too many sci-fi movies. AI is failing spectacularly at so many things. Yes it can produce text that looks like a human wrote it but it cannot be funny or have original thoughts and it also makes shit up all the time and no one knows how that can be prevented. It can generate images but there's a lot of randomness and you can never get exactly what you want, and again it can't generate anything original. Most people hate how AI is crammed into everything right now and I do not see that we would be putting an AI in charge of the electrical grid without any reliable way to manually override it. Like just what would be the point of that? AI is the hype thing right now and companies have noticed that advertising products as having AI will make stock prices go up. Eventually people will stop caring after seeing how poorly it works and we'll get on to the next hype thing. Sure, AI will replace some jobs and that's bad but trust me when I say that you will not be seeing any robots on the streets shouting "kill all humans" in your lifetime.
Is because AI is in their early development. And corporations don't need AI to br funny or imaginative or cleaver. Corporations only need to do the task asked. Also corporations are greedy, why pay a designer 100k a year, when they can pay a license fee for a fraction of the cost. People losing jobs because of AI is been already happening. Maybe it hasn't touch your industry or had an impact in your life. But in many sectors, the implementation has increased. And trough the years, that technology has improved.
Companies like Amazon are already heavily automated. There is sections where humans are not allowed. Because there is hundreds of robots sorting and moving products. We might not get robots walking in the streets like in the movies. But you bet Companies would expend thousands in automation and AI if it means saving billions in human labor.
People hating AI because is integrated in everything. Won't stop companies in doing it. If it means save money. How many people complain about automated machines when you call a company. Because affected human interaction. But every company still has it.
@@firenitros2043 I agree with all that you said. AI is taking away jobs although the results are not always great for the companies. My comment was mostly about the videos concern for some huge disaster and a possible AI takeover. AI will not suddenly become hostile and try to destroy humans any time soon. When it comes to art and entertainment AI also doesn't stand a chance against humans. If youtube gets flooded with 90% AI content people will just stop watching. AI can make nice looking images but it can't put together a coherent narrative. That might change some day but regardless AI has to be trained on human input to even function at all. You can't train an AI on AI output because it messes it up. So we will always need humans to create new stuff for the AI. Again, that might change with a new type of AI but we cannot assume that every technology just gets better and better forever. Things stagnate. Smartphones haven't gotten significantly better for a decade for instance, they still do the same things as in 2014 only with better performance. There are limits to technology. But yes, I do agree that companies are replacing people with AI and it is really bad.
