Instale o Steam
iniciar sessão
|
idioma
简体中文 (Chinês simplificado)
繁體中文 (Chinês tradicional)
日本語 (Japonês)
한국어 (Coreano)
ไทย (Tailandês)
Български (Búlgaro)
Čeština (Tcheco)
Dansk (Dinamarquês)
Deutsch (Alemão)
English (Inglês)
Español-España (Espanhol — Espanha)
Español-Latinoamérica (Espanhol — América Latina)
Ελληνικά (Grego)
Français (Francês)
Italiano (Italiano)
Bahasa Indonesia (Indonésio)
Magyar (Húngaro)
Nederlands (Holandês)
Norsk (Norueguês)
Polski (Polonês)
Português (Portugal)
Română (Romeno)
Русский (Russo)
Suomi (Finlandês)
Svenska (Sueco)
Türkçe (Turco)
Tiếng Việt (Vietnamita)
Українська (Ucraniano)
Relatar um problema com a tradução
Regras e diretrizes para discussões
Denunciar esta mensagem
But it's doing the opposite.
We already have AI art generators. We have AI text generators. There are also AI music generators in development.
What's interesting about the text generators is that once it has a grasp of language, well it's a hop skip and a jump away from learning programming languages. So at some point we will also have AI writing their own programs.
Vernor Vinge's technological singularity becomes a possibility at that point because what happens if an AI writes the program for another AI? And what happens if that new AI performs better than the original?
Also they could combine all of the above and have AI video game generators. Want a new game? Describe it in a text box and click the button.
And then a multi-billion dollar industry goes *poof*
It's a bit more complicated than just pressing a button, despite being a huge increase in capabilities that it provides, however, it's chaotic and non-specific. The a.i. also only has a very weak understanding of popular concepts. Anything that's niche or specific still requires a human and will until AGI comes around (which artificial general intelligence has its whole set of issues to be concerned about, which Robert Miles has a Youtube channel covering a significant amount of - and AGI is potentially comparable to artificially manufactured human-beings)
a.i. generators take the tedium out of the work process. The only artists they're putting out of a job are the ones who lack creativity and were just relying on the inabiliity of the general population to perform the machine's-work of getting dimensions & scale & form & such accurate.
Now a machine can help people get scale correct and make the profession of art entail spending more time making actual design & director decisions, rather than having to spend hours drawing lines in the correct shape when that's time that isn't getting spent on making design decisions.
The a.i. generators don't make design decisions and computers are still quite a ways away from even being capable of doing so, despite the rather impressive tricks that this new technology can do.
Try using it if you get a chance, you'll really get a feel for how it's simultaneously really enabling but also really limiting.
Now, there is a bridge to this which will make it happen sooner but it won't be fully autonomous... you get something in a machine that understands ...enough... to be able to do what your saying with guidance from human directors.
So image generators need art-directors to guide them.
Code generators need senior software developers to guide them.
And text generators need authors to guide them.
They can't do it on their own yet.
It won't be until AGI (artificial general intelligence) is achieved that all humans can be removed from the job processes entirely (but this is still an oversimplification).
Any living creature needs a purpose ...and as long as we're relevant to that purpose, we won't be disposable yet...
https://www.youtube.com/watch?v=IeWljQw3UgQ
https://www.youtube.com/watch?v=zkbPdEHEyEI
artists were already worthless.
I made a video with some of my AI-generated artwork. It all succubi girls though.
I wanted to see Judgement from Helltaker in a dark temple type environment and there wasn't enough of that on the Internet, so I had Stable Diffusion make some.
I would argue this is fairly niche.
https://youtu.be/FSVIGBdxYrs
Imagine Overwatch without....
Nevermind. I'll wait.
Also I need a rig with a good graphics card. I'm stuck doing CPU rendering on mine.
*plays piano slowly*
Imagine all the people...
...
...living life in peeeaaaceeeee
This video doesn't even rise to the level of a storyboard, as there's not even a story in this video (ie. story: this happened, then that happened, then this other related thing happened, then there's a big reveal, etc. etc. then a resolution, and the conclusion)
It's also not animated. The randomness in image-generation currently makes it too difficult for such systems to animate the images generated. Current generation doesn't have a conceptual understanding of objects, subjects, & attributes - it's all just kind of mixed together.
Maybe those characters are exactly what you want and that's enough but there's plenty of more specific things that people want which the A.I. just can't do yet - the best it can do in those scenarios is create background assets and reference images that a client passes on to a professional artist with some notes written by the client as to what they like about the reference image and what they want changed.
Keep in mind also that succubus isn't really that niche of a subject - there's thousands of images for it to train on for this subject, whereas something like a highly specific game character that doesn't get a lot of drawings of them made, is going to lack sufficient data for the A.I. to be trained to draw them.
At minimum, artists will still have to be payed to make what the a.i. generators can't because they lacked the data for them and no one is going to do this en mass just to train the machines but over time they'll become able to be trained off of new commissions that are being made for specific content - but this also relies on the content to be properly tagged otherwise, it doesn't learn what it needs to (and sometimes still doesn't because learning a concept abstractly by just witnessing examples rather than having a proper definition, is inherently prone to errors).
The only artists this is going to be putting out of work are the ones that had no talent in the first place - or maybe that were really good at just 1 thing that is high in demand.
I'm not. Currently people aren't taking A.I. safety seriously enough.
https://www.youtube.com/watch?v=9i1WlcCudpU
I fully expect them to be implemented with a reward / high-score function guiding their actions, and to be prone to the A.I. stop button problem. Combine these 2 aspects in an intimacy robot and you get mechanical succubi. Robots that do everything they can to seduce you but won't let you opt out of engaging them in intimate acts ...forever... everyone in the world.
https://www.youtube.com/watch?v=3TYT1QfdfsM
https://www.youtube.com/watch?v=tcdVC4e6EV4
I mean... "death by snu snu" with something that's actually doing a decent job of trying to be attractive to me, doesn't sound like the worst way to go but it's also still... wrong... and I'm sure that most people wouldn't enjoy this, even if their body is betraying their mind.
Men don't want it nearly as much as they're stereotyped to...
This would make an awesome hentai btw.