A lot of pressure, thinking that the road is gone before getting on the bus?
But as I read a lot of posts about AI painting and a lot of analysis videos of big bosses, I gave a simple attempt. To sum up, in my personal opinion, it is still too early for AI painting to replace painters, at least in a short time (two or three years). After all, it will take some time for a thing to go from perfection to mature commercial use, but there are still many shortcomings in the current AI output quality, such as lack of clarity, similar style and imperfect details (AI can't draw hands, hahaha!
) and so on, these shortcomings are obviously that AI is drawing. At the same time, it seems that most AI is just a "tailor", not a "painter". AI is just tinkering on the basis of previous paintings. Personally, I feel that I have got a good painting with almost all keywords. At first glance, it was amazing, but when I calmed down, I liked it.
? Now, those AI paintings that are not bad are just high among many short people. After use, most of the paintings are invisible and have a lot of evil spirits.
(Maybe it's because the AI I use is too rubbish), which is actually a hundred thousand miles away from my requirements (it may be ok to meet personal needs, but can it really meet the demanding Party A? )。 At the same time, the most important and crucial issue is copyright. Now the paintings that AI uses for training are all from the Internet. If the painter doesn't pay attention, it may make AI cheaper. But since everyone has noticed the infringement of AI, I don't think many painters are willing to give their paintings to AI for training. I believe there is always a way to deal with the plagiarism of AI (maybe you won't see the high-definition pictures of your favorite artists in the future, or you may reduce the number of sharing, ho ho!
), at the same time, it can't solve the copyright problem, so it is unrealistic to use it for commercial purposes. Perhaps most of them can't define plagiarism, but there is always a possibility of a highly similar situation, which may lead to copyright disputes. AI that can't be used to make money is not a good AI(doge). And to be honest, I learned about AI painting in the first half of the year, but the development of AI in the past six months has not reached the height I expected at that time, so how long will it take for AI painting to develop and improve? I think everyone can remain optimistic.
However, the future is really unpredictable, and no one knows how far AI can develop.