- cross-posted to:
- imageai
- cross-posted to:
- imageai
The two obvious themes in my mind are gears/some sort of complex machine, or computer hardware/circuits.
As an avid Lain Iwakura enjoyer, I added that screenshot because it’s awesome and I think it’d be an extremely cool banner, but there are probably much cleaner designs out there that would be more welcoming to new users. I’m sure you guys will be able to come up with some in short order.
You can look on lemmyverse.net to get a quick overview of some other instance banners.
I really don’t know how the formatting and proportions are supposed to work but I will edit that in as soon as someone clarifies.
Lemmy.world looks like this
This discussion will run for at least a week, and then we can take the most upvoted submissions in this thread and put them up for an official vote.
OK, so here are a few suggestions from me as a member of the imageai community:
Lemming working on computer 1
Lemming working on computer 2
Happy tech seal
Love the first one
Agreed, first one has my vote.
The top one is my favorite but they’re all super good. Nice work
Not their work though, it was a collection of random artists work conglomerated by an ai which they then prompted until they got these.
Would you congratulate someone who commissioned art from an artist for their work while arguing it to be theirs? No, it might be theirs since they paid for it, but it’s not their work, even if they described and revised it. Here, no payment or agreement was made. This is just theft.
It’s artwork that was created by a human being operating a machine in order to achieve a certain aesthetic.
How is that any different than an artist using Adobe software to create digital art? They didn’t create the art, they simply paid to use a machine that created the art and then claimed that they made it.
In both scenarios, the art could not have been produced by the artist alone, nor could it have been produced by the machine alone. What’s the difference?
The technique, brush strokes, composition, colouration, medium, etc at every step is decided by the artist when they draw it. Each part is crafted.
When you prompt, you simply get a new image and keep prompting until those aspects are what you want. Using others technique, brush strokes, composition, coloration, medium, etc. While only adding concept. It is equivalent to asking an artist to draw you something, revising, then pretending it was your own work, not paying the artist, and doing so without their knowlage
If you have permission to use these things, there is no moral issue, but as it is these models are trained on all public data. From any artist, mostly smaller completely unpaid artists
The diffrence is with photoshop you make each decision. Any you don’t through filters or separate tools come with the permission and knowlage of their creator. They did create the art, with permission to use what they used, with full control over what makes the art art. It’s technique, brush strokes, composition, colouration, medium, etc. This past just describing it, it took skill.
I am a developer, I understand how these AI’s work, and have made tooling for them myself. No, prompting is not equivalent to the skill of drawing art. When you prompt
(in the style of davinci:1.3)
this in no way means that you understand what davinci’s style was, and even if you did describing that instead would be detrimental to your output. The AI was not trained on an accurate description of the concepts involved. Just “a painting of a woman sitting facing the viewer with a slight smile in the style of davinci”. Any more description would degrade the result.Thank you for the reasonable responce though
I don’t necessarily disagree. But I guess my feeling is that you can’t shut the stable door after the horse has already bolted. AI art software is already plentiful and widespread.
Within our capitalist system, AI art will be utilized because not using it is a competitive disadvantage. Many artists are going to lose their jobs to technology just like so many other fields already have. It is what it is.
This doesn’t make it moral, nor does this have to be the case. For an extreme example, slavery is always profitable. We see much less of it today through regulations than before. Thing is, we can. It’s not hard to tell what’s AI and what’s not with the right tools, and unless they can prove they have the rights to their models training content this should be banned. Heavy fines just like when they trace art today, or steal it in other more old fashioned ways. That too is more profitable, but rare due to regulation
AI art is new and without regulations as it is we will see the worst possible outcomes. This can be hampered with regulations or broad lack of support by people like us. We can simply reject it. If there is no demand supply will be useless.
Will this slow down progression? Yes. So has banning forced labour in amarica. China and russia had an absurdly quick leap forward heavily utilizing it, does this mean we should re-instate it? If we focus solely on progress at the expense of people we have nothing to progress for.
When mechanization took field work away mechanics jobs, office jobs, and service jobs came up. This took a hard laborious job which damaged a person and allowed people to pursue better jobs. AI only takes, with no benefit to the common person.
This will be beneficial in the long run, cutting this out now. AI is a useful tool but it can’t create new techniques or concepts. It needs input. If let to be how it’s going it’ll stagnate our culture. In capitalism stagnation is death.
We can’t feed ai output to an AI either so with the mass proliferation we’re getting close to the peak on development even using our current methods. Model collapse is much worse than we thought it would be and we have no solution. Newer models and models trained the same way today are markedly worse.
Opinions change on the scale of a person to another. The real way forward is through conversation and political engagement. People hate to hear it, but this does work. Look at the EU, with their recent acts to regulate AI. Things can be better.
We can track down the horse, we can build fences to stop it from getting too far next time, we can close the barn door to keep the other horses in, we can shoot it and use it for meat if it’s too quick. There are ways forward, we don’t have to just let this horse run wild.
You mean well and I see where you are coming from, but I personally find your mindset to be somewhat naive. I wish I believed that conversation and political engagement would somehow solve our problems, but its going to take much more than that. There is an ingrained worldview that is fundamentally flawed.
I agree that there is a way forward, but I sincerely doubt that it will come by working within the current political and economic system. It will be necessary for groups of people to escape the system entirely in order to gain leverage to actually change things.
I think that much of your wariness about AI art is only applicable within a late stage capitalist paradigm. If, for instance, artists didn’t have to sell their work to make a living, the downside becomes much less clear.
What would your suggestion be then? I’m not opposed to more radical plans, even [likely bannable] actions, in many circumstances specifically climate I see it as a neccesity, but I don’t see the case here. You may call it naive, but this has worked with other issues and will continue to work today.
As a socialist yes, I agree, the system needs radical change, but that cannot happen through violent action, look at all previous violent attempts. nor would I argue it to be needed here. Again, in Europe this is already happening within our flawed system.
And yes, under a system where peoples basic needs this wouldn’t be as immoral, but it’s still not art and is still theft.
To reiterate a person who prompts for the stle of ghibli does not need to know what that means. The absurd amount of effort that went into creating it, or the time it takes to hone that craft. In fact, if you were to describe these things, AI could not give you something coming close to the quality of “in the style of ghibli”. This is a fundamental issue too, not something that will get better with time
I like #2 myself, he looks happier, and cuter. #1 looks a little sad.
#1 is doing all the actual work. #2 is his manager.
Any chance you could make a few more with a prompt like “Satisfied seal working on computer”? As the satisfied seal is our logo, would make sense to work it into the banner too.
I’ll give it a try, but seals in Midjourney aren’t as… pliable as lemmings, if that makes sense. They can’t really work with their flippers and are more closely associated with certain styles that the AI insists on inserting.
I’m going to have to be strict with it. I’ll post new seal based ones if I’m satisfied with the results.
Ah, fair enough. If it’s not a thing, it’s not a thing. I appreciate you giving it a shot though :)
I love number one for the disorganization
All three are good, but I really like the top one as well. I just wish it was a seal
I like #1 best, but all these are great!