About as skilled as "google searching" - some people are very bad at it. but most are good enough.
Google searching is a skill, and most people are pretty bad at it.
Quite different things - it is a false equivalence. If you are writing your own texts and translating them, they are still unique and based on your experience.
It is not false equivalence. You have no skill in translation, no experience in translation. So the text you get from an MTL has no basis on your experience or your skill. It might be unique, and so is art generated by AI. AI-generated art is basically a translation of words into images. It is the same thing, just the output of the medium changes.
No. People knew they were coming and have known for many years. People hate on them because they are unskilled, but mimic skills. It isn't like a calculator that makes mathematical equations easier for those who understand the math to begin with.
No people did not know AI-generated artworks were coming, for years they thought art would be the last bastion to be untouched by AI.
It is a false equivalence, because you actually fed it something you created to translate. It isn't plagiarism. Even if paying a translator, it still isn't plagiarism. However, paying an artist to paint a picture and then say you painted it, would be unacceptable, right?
Not sure what your experience is, but this is not the case.
I must say, all this AI came in like a lightning strike for me. I'm on a computer all day every day, with large portions of that time watching youtube, news and other content creation based around sciences and generally on the more academic side.
I've obviously been aware of AI being a thing and the philosophical discussions around it - music AI has been around for years - but the suddenness of the implementation of AI art and then, out of nowhere, ChatGPT and that coding one that can now do high level coding... it seems to have been blasted out of a cannon into my face.
In this regard i'd consider myself fairly representative of your average person, a tinge more connected to frontier developments than most. So if its hitting me like that, i'm pretty sure it's even crazier for the majority of people
Perhaps it depends on what people are into. I think that for many of the people I have talked with, it is of little surprise, because anything that is codifiable, can be encoded into an AI. This is codifiable and photo editing software has been doing components of it for decades, it is just that people weren't paying attention.
The coding software has been available for a few years already to the point that many low-level coders aren't required, and again, this has been happening for a while.
But perhaps for those who are paying attention to the current outcome, aren't necessarily putting all the pieces together that lead to what is not too far down the track.
If you paid a translator and said you translated it. It would be unacceptable, right?
Also, translations are protected under their own rights, so yes it is plagiarism.
It has been the case for many years until diffusion models come along.
AI still hasn't generated art, other than the art of the code that leads to the generation. It has generated an output that mimics the art, but doesn't have any journey of the artist required to make it actual art. I can take a photo of a tree, but my photo doesn't live the life of a tree.
And do you think machine translators lived a life of a translator to create translations? Did MTLs learn a language's intricacies? Did MTLs spend hundreds of hours translating documents, poems or novels? You know there is a reason there are multiple translations for poems because translation is subject to that translator's own journey.
Translations are not different from art. So why MTLs are acceptable but AI art not?
You have still missed the point of the false equivalence. A translation is based on an actual piece of writing by a person (if not AI generated) who had a journey. An AI generated image is not the same.
For example, if you grab a pen and paper and draw some Manga from your own hand and mind, how close do you get to the images you generated on your blog?
As I told you there is no false equivalence.
The person who has used MTL has no skill at translation, or do you think they could get close to the translation the machine spit out? If they used their own hand and mind?
And those images on my blog are still based on a piece of writing from a person. ie the prompt and all the other options like CFG Scale etc.
And I have proven that your claim that this is false equivalency is not true multiple times. What you are doing is just proof by assertion and if we are not going to argue in good faith there is nothing more to add.
No - you haven't proven anything. And, by the lack of answering the actual question I asked, I will assume you can't draw.
It came up slowly though...it exploded this year...no body saw that coming.