Just griping.
I was asked to remove or hide behind a content warning any AI art I feel like posting to the Mastodon instance (Elekk.xyz) where I made my home when it looked like Musk was going to turn Twitter into his own personal fiefdom. That’s looking increasingly unlikely, but I have met some cool folks on Mastodon, so I still use it. Sometimes I post there things I don’t think would fit somewhere else — like my friendly wombat, here.
This isn’t the first time I’ve been called to the principal’s office. When I first joined, I did the thing where I would “friend” anyone who looked fun or interesting. I was told I shouldn’t friend anyone unless we had interacted and they seemed agreeable to the prospect, as being suddenly friended was triggering to people.
Okay, I won’t friend anyone without discussion. Your site, your rules.
Next, I was reminded to always use ALT text on images I post so that those with sight issues can get the meaning from the image. Actually, yes, I should do that, and have been mostly doing that both on Twitter and Elekk.xyz ever since.
Now I’ve been told that posting AI-generated art triggers people with disassociative conditions, as they look a little bit “off”. So I went through my timeline and deleted any AI-generated art — including the Wombat — that I’d posted, and promised to hide any future posts behind a content warning that blocks the post from view unless someone clicks the warning and chooses to see it. Of course, nobody is going to bother, because it could be okay, could be horrible, no way of knowing. So I simply won’t post that sort of thing there anymore.
Question is — am I the baddie? I’ve been posting AI stuff all over the place for a couple months now. I still plan to post to Twitter — if people don’t want this, they can just block me or simply not follow me in the first place. I dunno.
Right now I am a little mad that it looks like Elekk.xyz won’t be a permanent home in the Mastodon Fediverse, and also a little mad that someone who was struggling with life in general saw something I posted and was made to feel bad.
This isn’t an AITA. Not looking to bash the Fediverse in general or Elekk.xyz in particular. I just wanna post AI stuff for awhile.
So, fair warning — if you come to this blog, I can’t promise you won’t see AI generated imagery. But, I’ll try to keep it light.
All I can promise at this point, is more wombats.
I’d be curious to know if it’s possible for people other than trained art professionals or similarly experienced enthusiasts to differentiate easily between human-generated and AI-generated art in similar styles. A lot of pre-digital art looks “off” to many, many people, which is probably why so much of it has been so unpopular and controversial over the years.
On the issues of accessibilty and inclusion, at what point does ensuring no-one feels left out morph into making it unlikely anyone will be left in? Personally, I am fascinated by AI generated art, writing, whatever. I suspect that trying to avoid it in the future is going to be likle trying to avoid mobile phones, television and advertising in previous eras. Plenty of people wanted to but most people either accepted it or embraced it because they really didn’t have a choice.
There’s also the issue of user-responsiblilty as well as producer. Nothing you’re going to produce is pushed to the user. The user has to pullit from you. So long as you flag it up, I vote for as much AI-generated content as you want to give us.
The “Fediverse” has the concept of a local feed, which is a feed of everything that is posted on that instance, so in that respect, I am forcing people to view stuff they haven’t asked to see.
I’ve been thinking about this all night. Not about whether or not to continue with my AI project, because I’m gonna. But I wondered how it would feel to be in such a shaky mental situation that I am constantly being pushed over the edge by ordinary situations, and yet I would continually expose myself to the possibility of seeing things I would not expect.
As I’m writing this, “Thirteen Step Program” has just been auto-published, and this is a post where I felt the art DALL-E 2 generated was genuinely disturbing — to me, who specifically crafted the prompts that caused it to be generated. So I am not entirely unsympathetic to this. As this month goes on, I feel I’m getting a better understanding of how AI art and text intersect with human creativity, and how a dispassionate machine perspective can take me to unexpected places.
Oh yeah, re: can people tell between AI and non AI art, an artist accused Midjourney of stealing her painting style without permission. Midjourney showed that the particular prompt (“Portrait of a Woman”, I think it was) produced the same output whether or not it was specifically prompted “in the style of this artist”, and in fact couldn’t find her name in the data. So maybe even artists can’t really tell — or their art isn’t as unique as they’d thought.
Mastodon (or at least that elekk.xyz stream?) seems like it might be a rule-lawyers dream. For anyone else… Heh.
I’m yet to be bothered by anything enough to take a serious look at these federated social media platforms. I barely keep up with the ones I have, let alone digging down into the microcosm of mini-platforms.
As you say, their site – their rules, so whatever. But it seems to me to just reinforce my existing view that it isn’t worth it.
I do hope you’ll continue posting to Twitter though, definitely no blocks from me! 😉
If I didn’t post to Twitter, hardly anyone would read my blog. Most of my referrals come from Twitter!