YSK: db0@lemmy.dbzer0.com set up a AI image generator bot on the Threadiverse for anyone here to freely use
tal @ tal @lemmy.today Posts 157Comments 6,642Joined 2 yr. ago
tal @ tal @lemmy.today
Posts
157
Comments
6,642
Joined
2 yr. ago
I think that at least some instances use Cloudflare for various things, so depending upon what and how much stuff at Cloudflare is broken, some lemmy instances may be impacted.
When you stop using a search engine and use AI-generated responses for everything, I guess having your LLM backends go down could qualify as having a disruptive functional impact.
Google Cloud status page:
The reason many AI image generators do multiple images is as a simple way to trade compute cycles for quality. The idea is that you generate a couple and pick the best, using your human knowledge of what you intend.
You could generate it in one place, copy the URL of the best image, and embed it in your response. That's what I did when I pasted the links to the skunk engraving images in my post; the images were generated elsewhere. I just pasted all four rather than only one, to show what the response looks like.
The syntax for an inline image on the Threadiverse's Markdown variant is:
I assume that either that syntax or a similar one will work on Mastodon, but I don't know Mastodon's syntax, as I don't use it.
I put this in the post body, but it was further in and I think that some people may not have read that far: this community, !YouShouldKnow@lemmy.world, says that it bans most bots, so I don't think that that bot will operate here. I linked to a test post on another community, !test@sh.itjust.works, where I know it works, because I just tested it there, if someone just wants to give it a try.
https://lemmyverse.link/sh.itjust.works/post/40008132
or
https://sh.itjust.works/post/40008132, if your client can't understand the above.
If you only care about folding Markdown, as in your animation, I'm sure that there are Markdown-specific editors that do that.
If you want folding across a wide variety of languages, then I think that your choices are going to be more-limited, since those editors need to be able to parse those languages. Vim and Emacs are kinda the Big Two general-purpose editors, and they're gonna have the widest support.
EDIT: For specifically programming languages, a number of IDEs can probably do it.
Here's Eclipse, for example.
EDIT2:
https://en.wikipedia.org/wiki/Comparison_of_text_editors
This has a "text folding" and "code folding" column.
There are ISO country codes.
https://en.wikipedia.org/wiki/List_of_ISO_3166_country_codes
Most AI image generators that generate images add EXIF metadata indicating that the image is AI-generated. This helps people who want to identify AI-generated images readily.
In the case of ComfyUI, it even includes the entire workflow --- like, another ComfyUI user can just grab the image, drop it onto their ComfyUI Web UI and they'll be right where the generating user was.
Unfortunately, EXIF metadata can contain location information --- some cameras and such add it --- and this metadata led to people posting images at places like Reddit being doxxed after they didn't realize that they were posting their GPS location and maybe real name, stuff that some cameras attach. As a result, a number of image-hosting places simply strip all metadata, to prevent users from from accidentally leaking this information.
Pict-rs, the software package that Lemmy hosts run to permit image uploads, does this. Unfortunately, it means that those "this is an AI-generated image" tags get stripped off.
So, for example, on my system, with ComfyUI, using ImageMagick:
$ identify -verbose output/ComfyUI_00312_.png
"Properties:prompt" has a JSON encoding of the workflow.
Sample images generated by various AI image generators are readily-available on civitai.com.
For this generator that generated this image on civitai, it looks like the parameter is "Properties:parameters".
I believe that there are a small number of such tags today.
It would be technically possible to just not have pict-rs strip that particular tag (or tags, if there are multiple that a given generator adds?) off, have a list of "AI-generated tags", then have Lemmy add some visual indicator that an image is AI-generated. I'd suggest that this is probably a better longer-term route to indicate that an image is AI-generated than manually-tagging post titles, for a couple of reasons:
- Spiders that index images on the Web will know that the image is AI-generated and can flag that for users and let them use that as a filtering criteria (e.g. Kagi Images permits for this). They aren't going to understand tags in post titles, but the metadata tags are somewhat universal.
- Doesn't require manual effort if an image can have some indicator or flair or whatever put on it automatically. And I guarantee that some users are going to get this wrong just by accident, because different instances have different rules. Easier to change how a computer works than to change human behavior across-the-board.
- Works on all instances.
- The information remains attached to the image even if downloaded.
- Works for images that aren't just the subject of single-image posts and don't have an associated title.
- Speaking purely for myself, I kind of like the open-source, collaborative aspect of sharing the workflows or prompts, since it helps other users see how an image was created and learn from it; it's something that I'm glad to see the generators include, and I'm kind of sad that we strip it off on the Threadiverse.
Yeah, I remember reading about AI Horde a while back and thought about it, but wasn't comfortable with having an system that fields requests from the outside world, as my GPU is on a desktop and I'm not totally sold on how well these guys have hardened the thing. If I had a dedicated compute node for doing renders, I'd probably be fine with it, as it just eats some electricity then.
Hahaha, that's fantastic. Thanks, @db@lemmy.dbzer0.com. I can't believe that I had been unaware of this for years.
I've got local hardware myself, but this makes generation way more accessible to random users on the Threadiverse without tying them to commercial services, including users who are using a phone and can't really run the stuff locally.
skims bot FAQ more
It does look like the bot tries to block NSFW stuff, but outside of that, it looks pretty permissive of whatever, and while the style list doesn't clearly distinguish between styles consisting of prompts for one model and different models, there are a lot of models in there that I recognize. Just append style: flux to the end of the prompt. It's got Flux (style: flux), Pony (style: pony), Pony Realism (style: pony realism), the Nova family (style: nova anime xl, style: nova, style: nova furry pony), Stable Diffusion 3 (style: sd3), SDXL (style: sdxl), Mistoon (style: mistoon anime), just from a very quick skim. I guess AI Horde must have a list of models and the ability to just download models or something if it doesn't already have them present on a node?
skims bot FAQ
Oh, dang, it even runs multiple models, and it looks like it can run Flux, which can handle fairly English-language-looking prompts.
I gotta see this.
@aihorde@lemmy.dbzer0.com draw for me An engraving of a skunk. style: flux
Hahaha, awesome, db0, didn't know that you'd set this up until now.
1986 – The Mindbender derails and kills three riders at the Fantasyland (known today as Galaxyland) indoor amusement park at West Edmonton Mall in Edmonton, Alberta.[10]
Huh.
https://en.wikipedia.org/wiki/Mindbender_%28Galaxyland%29
On January 30, 2023, the mall decommissioned and closed the Mindbender after 37 years of service, in order to redevelop its space for new developments in the park.[4] Its trains were reused for All American Triple Loop, at Indiana Beach, United States.
https://en.wikipedia.org/wiki/All_American_Triple_Loop
All American Triple Loop (formerly Montaña Infinitum ["Infinity Mountain"] (2007–2014), Montaña Triple Loop ["Triple Loop Mountain"] (2014–2016) and Quimera[1] ["Chimera"][2] (2017–2019)) is a steel roller coaster at Indiana Beach in Monticello, Indiana.
Manufactured by Anton Schwarzkopf, it was originally purchased by showman Rudolf Barth in 1984 who operated it as Dreier Looping for 12 years on the German fair circuit.
After this, it was the main attraction in three major theme parks: first spending 2 years in Sunway Lagoon as Triple Loop Coaster, next, it spent 5 years in Flamingo Land resort as Magnum Force, and finally at its third and most recent location at La Feria Chapultepec Mágico, as Montaña Triple Loop. In 2017 it was renamed Quimera. In 2024, it opened at Indiana Beach as All American Triple Loop.
I feel like a secondhand German roller coaster that went from Germany to Malaysia to England to Mexico now running secondhand Canadian trains arguably isn't best named the "All American Triple Loop".
The review to determine whether the US should scrap the project is being led by Elbridge Colby, a top defence department official who previously expressed scepticism about Aukus, according to six people familiar with the matter.
Eh. Looking back, it sounds like his take is fairly nuanced. From what he's said in the past, it sounds like he's said that he doesn't think that entering the arrangement was actually worthwhile for the US, but that he's also hesitant to withdraw from an agreement once entered into.
It sounds like his argument is that the main risk of a military conflict with China over Taiwan, where these would play a role, is relatively near-term. Australia hasn't stated that it would defend Taiwan, and AUKUS won't result in an aggregate increase in submarines across the US and Australia for some time, which means that it would reduce the number of submarines available to fight China.
Assuming that all that is accurate, that seems to be a fair take to me. My guess is that what he's actually after, given his phrasing, is not trying to trying to end AUKUS, but to get Australia to also commit to defending Taiwan as a condition for it.
Speaking in London on Monday, Mr Colby said that US shipbuilding could not keep pace with the target of delivering Australia subs by 2032 and questioned why the US was giving away its most lethal assets to a country that was not even guaranteeing it would use them in the event of a conflict over Taiwan.
“If I were king for a day on the subject I would say ‘Look, you all know what my concerns are, let’s see if we can work through these together’.
I've no idea whether that's something that would be totally unacceptable to Australia or not.
At one point, the US did try to put together an analog to NATO in the Pacific, SEATO. It didn't really go anywhere. But conditions have also changed.
https://en.wikipedia.org/wiki/Southeast_Asia_Treaty_Organization
Australia was a member. Taiwan was not. But the US might be aiming to build a new Pacific alliance today.
Deleted
Permanently Deleted
Jump
I think that that's pretty much a subjective question. I could tell you what I find tacky, but it won't be a universal. Some people will have different positions, and there isn't one "correct" view to have. It's like asking what the best flavor of ice cream is, or the best song.
https://en.wikipedia.org/wiki/White_House_Rose_Garden
I think that's gravel, looking closely at the image.
That being said, I don't know whether concrete walkways were much of a thing in 1908. Gravel might have been the closest common option to what we'd more-commonly use concrete for today.
Not as big a deal in DC's climate, but here in often-arid California, it's generally considered responsible to phase out water-hungry lawns in favor of landscaping that doesn't require as much water. Drought-tolerant plants, gravel, rock gardens, concrete, whatever.
https://en.wikipedia.org/wiki/Xeriscaping
The lawn was an English custom, and trying to reproduce a little piece of wet England by pouring enough water on arid or semi-arid land day in and day out is kind of wasteful.
If you need the cushiony and rapidly-self-healing properties of lawn because people are running around on it, that's one thing, but people spend more time indoors than they did historically, and as just a thing to look at, it's not a great default. Plus, kind of high maintenance.
A factor for a number of Western states.
That being said, this wasn't the rationale, and frankly, it's probably basically irrelevant for somewhere like the White House relative to the functional impact.
A couple of my favorites from that thread:
@Usernameblankface@lemmy.world (also OP here, also the person who brought the idea up in another thread)
Link to the original version of this thread to which OP is referring, which was quite a hit some time back:
Yeah, just one word to refer to all the compatible Fediverse Reddit-alikes.
Sublinks will be in there too, if they get the ball rolling.