I think what you’re missing is scarcity and I think it was a problem even before AI. When everything was harder and more expensive only the best was able to get made and there was less to pick from. That actually resulted in scarce amazing music, movies, art, etc. Better software, faster computers, were already bringing the cost/barrier down. AI just basically takes it to near zero. The flip side of the downside is your wildest ambition is probably a totally doable thing now where it would have been impossible before, but that’s true for everyone now so, yea, not as special.
Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.
Yea, this kind of stuff is wild. I just want to make cool stuff and have fun. I'd rather AI not end humanity please. I think there needs to be sensible regulation. I'm honestly terrified of the local models I have running on my PC. There are no guardrails essentially and there needs to be some protection for regular people. With great power comes great responsibility and all that.
People will of course try it, but I think it will fail. There are movies that people spend millions on that use no AI that fail all the time. I can't believe that in an environment where that is true, total AI end to end will succeed. It's just hype, it's marketing. This is what I think.
While that's possible, I don't think it's financially viable. People appreciate art precisely because a person made it. We're still in the hype phase of AI, but I don't think full AI end to end has much viability. It's not visually compelling usually, it's not narratively compelling. It's not really that interesting from an art perspective. It might be cool as a tech demo, but not as a human expression which is what everyone wants from art.
The copyright issue is an issue where I think it makes sense. You write a prompt, the end, that is the final output. There is no expression in that, but I think what's interesting is using AI in a way that it would be pretty undeniable that a human is modifying and changing the output. This is where the potential is, not write prompt, have zero input in what the AI makes. That really is slop.
It's just another tool. Why use Substance painter if you can paint all your textures by hand. Why use Houdini for procedural setups when you could do it the long way. Why use 3D animation over cell animation. It's the same calculus, but AI makes it a bit more complex because it can go all the way to the artist being completely removed. The fact that that's an option doesn't mean it's the best option or what you have to do. There's a whole spectrum of use cases and applications.
Supply and demand goes both ways. Cost of production goes down, it doesn't just mean everyone gets fired, it also means more stuff gets made. Things that didn't make sense before, suddenly make sense. Games used to mean console, or PC. Then there was mobile and social. People might think a lot of those games suck, but all those mobile and social game companies = more jobs for artists. I can see a situation where AI shows/movies find a market and end up meaning more jobs for artists at the entry level. It's not all doom and gloom.
This is very true and also an issue that comes before AI. Streaming already created a situation with oversupply, it's true, but there's also the situation that things that wouldn't have gotten made, were able to get made. I like the idea of more people having access to making things. Will that open the door to a lot of crap? Yes. Just like Youtube and Tik Tok opened up the door to brainrot and a ton of content that maybe doesn't need to exist. There's also a lot of success stories in that.
Here is my example and maybe it'll make more sense with a specific example. I wanted to make a series with realistic human characters. Dozens of sets, dozens of characters, speaking, acting, a full on thing similar to a TV show. My background is in 3D and 2D art. Making a single full 3D set is weeks of work. Making dozens becomes unrealistic. Creating realistic talking characters basically takes a whole studio, but with AI I created what I wanted as a single person sitting at home. This is a situation that would be impossible with AI. I didn't do it out of fear, I did it because it was insanely fun. I used AI to design, but I also used my own designs. Watching them come to life right after I designed them is honestly like magic.
Well, of course there is. Again, you don't HAVE to do it. I think that narrative is silly, the whole "You'll get left behind if you don't use AI" That's pushed by people selling AI. It's an option. Just like you mentioned, using puppet animation is an option. AI is just another option.