Traditional_Land3933

joined 1 year ago
[–] Traditional_Land3933@alien.top 1 points 11 months ago

The sad truth is that even if such a thing as AGI can ever exist, we won't ever see it in any of our lifetimes, it's prob the conversation for many decades and centuries in the future, and where we are now is pretty much the nascency of widespread, conscious AI use among the masses. Of course, we've all been using AI for years but with these GPT-connected chatbots it's become much more common and active a decision to use AI than it ever was before. ChatGPT is the tip of the iceberg, we don't even know if the true best model going forward is going to remain the transformer (and it's not the best at everything, just the most "generalizable" at the moment as far as I'm aware). What I'm wondering is how advanced the societies of the distant future will be, where they look down on us and the state of our relatively primitive AI as we do monkeys with their great stone nutcrackers

 

I was explaining some of the stuff I work with in class to a friend who wants to be a career artist. She knows it's a hard career to make it in, but she's got talent and believes she can do it, so who am I or anyone to say she can't? Anyway, I was talking about GANs because I had an internship where we were working with them for a project. When I described what they were and what they were designed for, she started to get a bit concerned. It wasn't that she suddenly doubted her own talent, but when it comes to AI we don't know where the upper limits are (if there even are any), so in theory down the road they can present legitimate challenges to working artists. Not only would the requisite talent to work in the arts go up, but if AI gets good enough, in theory within the next few decades generated art can be genuinely good and tough to distinguish from high tier human work. I'm still a student, I don't know much about discriminators or AI detection, and it's not something I look into much in my free time. Is this a legitimate fear for artists to have? Or are there techniques which can pretty accurately identify GAN generated work most of the time?

 

I was never much of a basketball guy, so I don't know the ins and outs, but I'm trying to learn. When I ask this question, I mean with regard to which is the correct choice. Ultimately when you're in the zone you won't necessarily have a conscious, in depth thought process about everything you do, but from an objective standpoint, which is better, to attempt a dunk or attempt a layup? Obviously there's gonna be situations where one's easier to put in than the other, but assuming that somehow both are about equally open, maybe imagine a fast break scenario. And I'm talking about, like, for an NBA pkayer who's gonna have practiced both of them a million times and have the requisite athletic ability to achieve either of them more often than not.

This question came to me because, in football we teach not to try and extend the ball across the goalline, especially when diving towards a pylon. You see it done all the time in the NFL anyway, but there are times when it costs guys like Justin Jefferson a few weeks ago against the Eagles. There's a high probability of scoring anyway when you do it which is why it's done so often, but the chance of giving up a near TD or near goalline downing of the ball for a touchback is real enough that we teach not to do it. Now for a layup vs dunk, both are presumably the two best shots you can take in a high level basketball game anyway, but which is better? Dunks are literally placing the ball directly into the basket, but carry their own risks and can be tough to do even for guys that tall with that high a jumper and that much athleticism. Whereas a layup seems safer and more consistent but it's also not as direct as literally putting the ball in there with your hand and ostensibly (to my untrained ass eyes) seems easier to block than a dunk from someone as strong as high level basketball players are