Adobe deletes Bluesky posts after backlash

Adobe is the one major company trying to be ethical with its AI training data and no one seems to even care. The AI features in Photoshop are the best around in my experience and come in handy constantly for all sorts of touchup work.

Anyway I don't really think they deserve a lot of the hate they get, but I do hope this encourages development of viable alternatives to their products. Photoshop is still pretty much peerless. Illustrator has a ton of competitors catching up. After Effects and Premiere for video editing are getting overtaken by Davinci Resolve -- though for motion graphics it is still hard to beat After Effects. Though I do love that Adobe simply uses JavaScript for its expression and scripting language.

Adobe isn't trying to be ethical, they are trying to be more legally compliant, because they see that as a market opportunity. Otoh, artists complain about legal compliance of AIs not because that is what they care about, but because they see that as their only possible redress against a phenomenon they find distasteful. A legal reality where you can only train AI on content you've licensed would be the worst for everybody bar massive companies, legacy artists included.

> Adobe isn't trying to be ethical, they are trying to be more legally compliant

Is the implication of this statement that using AI for image editing and creation is inherently unethical?

Is that really how people feel?

For creation, yes, because of the provenance of the training data that got us here. It was acquired unethically in the overwhelming majority of cases. Using models derived from that training is laundering and anonymizing the existing creativity of other humans and then still staking the claim "I made this", like the stick figure comic. It's ghoulish.

There exist image generation models that were trained on purely licensed content, e.g. Getty’s. I don’t know about Adobe’s specifically-but if not, it seems like a problem Adobe could easily fix-either buy/license a stock image library for AI training (maybe they already have one), and use that to train their own model-or else license someone else’s model e.g. Getty’s

They are training using licensed images! That’s the thing! There’s some sort of ridiculous brainworm infecting certain online groups that has them believing that stealing content is inherent in using generative AI.

I watch this all quite closely, and It’s chronically online, anime / fursona profile picture, artists.

Exact same thing happened when that ‘open’ trust and safety platform was announced a few months ago, which used “AI” in its marketing material. This exact same group of people—not even remotely the target audience for this B2B T&S product—absolutely lost it on Bluesky. “We don’t want AI everywhere!” “You’re taking the humanity out of everything!” “This is so unethical!” When you tell them that machine learning has been used in content moderation for decades, they won’t have a bar of it. Nor when you explain that T&S AI isn’t generative and almost certainly isn’t using “stolen” data. I had countless people legitimately say that having humans have to sift through gore and CSAM is a Good Thing because it gives them jobs, which AI is taking away.

It’s all the same sort of online presence. Anime profile picture, Ko-fi in bio, “minors dni”, talking about not getting “commissions” anymore. It genuinely feels like a psy-op / false flag operation or something.