top 78 comments

sorted by: hot top controversial new old
[–] 130 points 3 months ago (1 child)

This was the discussion on a recent post, which seems to be deleted now, about "AI" art on beer cans in the Mildly Infuriating community.

It's really upsetting how much art is now seen as AI looking when AI art only looks that way cause it was trained on great art. Anything stylized, slightly sepia, realistically shaded, etc etc, automatically assumed to be AI when some people have trained decades to develop this style themselves, but now it's just slop style

  • source
  • hideshow 1 child comment
  • [–] 57 points 3 months ago (3 children)

    Artist self insert, given Lemmy even accused them of being AI despite them documenting everything. That's unfortunately the only way to validate at the moment, although if I'm being honest I kinda dig behind the scenes. It is especially interesting if you can see a beautiful photograph and all the work they put into getting the lighting just right.

  • source
  • hideshow 3 child comments
  • [–] 29 points 3 months ago (2 children)

    Wait, is it inherently bad for an artist to self-insert? Isn't that one of the key ways they share their perspective? If there's some bizarre strawman-ing or self-aggrandizing going on it can problematic but the general idea of an artist self-inserting themselves or their ideas into their work seems fine to me.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 27 points 3 months ago (1 child)

    I don't know if the person you replied to implied it was a bad thing, but it's been a thing for centuries, Caravaggio self inserted himself and people from the street in his paintings, van Eyck hid himself in his paintings, Louis David painted himself among the Coronation of Napoleon, etc

  • source
  • parent
  • hideshow 1 child comment
  • [–] 52 points 3 months ago (1 child)
  • [–] 42 points 3 months ago

    This artist has it rough in particular.

  • source
  • [–] 97 points 3 months ago* (38 children)

    Honestly this is one of the biggest issues with it all that you don't really see talked about much

    Artists sink a load of time into a piece of art they are proud of, and because AI has got to the point now it's not always obvious at first glance to average Joe, so people rely on heuristics to spot it instead.

    For this artist apparently it's "yellows"

    For anyone who takes photos (or even video now) even half seriously, photos with Bokeh and good lighting get called out now

    You like making instrumental electronic or sampled music? Get ready for everything to be called AI

    Even writing: I always used to like using em dashes—they're a pretty versatile piece of punctuation for any casual speech style of writing. If you use an em dash now though, someone just assumes the text is AI. So I've generally just stopped; not because I want to, because AI and people (somewhat understandably) not having a better way to identify AI writing.

    Hell, I made an excel sheet the other day to work some stuff out and someone asked how I got the AI to do it.....

    Now I'm not saying it's everyone, but I'm seeing a group of people forming, who basically say anything they can't immediately understand how to do themselves must be AI. I have a suspicion it's the same demographic that would say "well I could do that" in a modern art museum or "all they do is press play" when looking at any kind of electronic music production.

  • source
  • hideshow 38 child comments
  • [+] 33 points 3 months ago (8 children)
  • [–] 14 points 3 months ago* (5 children)

    One thing that always makes me think something is AI written is the repeated use of sentences like "It's not X, its Y".

  • source
  • parent
  • hideshow 5 child comments
  • [–] 5 points 3 months ago*

    You’re absolutely right! Let’s see if we can find better ways to determine if something is AI-generated:

    • what other styles are giveaways?
    • maybe a combination of em dashes and bulleted lists?
    • I’m running out of ideas for this joke.
  • source
  • parent
  • [–] 2 points 3 months ago

    After several years of Word automatically changing a dash between two words to a wider dash (maybe not the same type as emdash?) it's weird to see those same dashes becoming a tell of inauthentic.

    When are we going to see the AI using two spaces after a full stop to try and pretend that it types like someone more than 50 years old?

  • source
  • parent
  • [–] 8 points 3 months ago* (last edited 3 months ago) (1 child)

    With photography it's also a grey area. I'm a professional underwater photographer. I've got thousands of dollars of specialized housings, lighting, wet lenses, etc, but the color is still wrong because water. Meanwhile some open water student swam by 40 minutes ago and made the whole area hazy.

    So for decades I've done various degrees of touch up and and color grading because what you see with your eyes looks nothing like what comes out of the camera because human eyes and brains do a lot of processing that camera sensors don't.

    And then there's generative stuff that I don't think impact the art of photography. If someone uses generative noise reduction I don't think that's a problem. The non-generative solution to that is spending an extra 10 grand on a better camera and lens, so it's not as much of a skill issue as a money one. I'm absolutely fine with people using generative tools for that stuff.

    So where's the border between acceptable and unacceptable post-processing? There's clearly bad stuff like wholesale fakes and such, but I don't know if we have a clear line.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 5 points 3 months ago*

    Smartphone cameras are almost certainly the largest source of photography in society today, and those things default to doing all kinds of automatic processing on the images they take.

  • source
  • parent
  • [–] 15 points 3 months ago* (4 children)

    I'm currently writing a book and I'm terrified of using EM dashes, even where they would be perfectly applicable. I always end up restructuring the sentences to avoid it. The only instance i keep it is for dialogoue interruptions like:

    "I bend over and spread my—"

  • source
  • parent
  • hideshow 4 child comments
  • [–] 12 points 3 months ago (19 children)

    I think what makes the em dash seem AI-like is... there isnt an em dash button on the keyboard.

    So typically the only way those show up is when markdown-to-blog-post tools convert normal dashes into stylized em dashes as a transpiling process.

    Stuff like hugo and etc largely do this automatically.

    So if a poster uses an em dash out of that context, it right away throws up a red flag cuz... howd you do that easily?

    Did you memorize the alt code to make one?

  • source
  • parent
  • hideshow 19 child comments
  • [–] 8 points 3 months ago (1 child)

    So if a poster uses an em dash out of that context, it right away throws up a red flag cuz... howd you do that easily?

    Did you memorize the alt code to make one?

    On KDE, the default keyboard layout makes it pretty easy to type the various dashes and quotation marks, so that's one way. Ever since discovering that, I've been making a conscious effort to use them (because I am the type of person who likes doing the typograhically correct thing, even in an informal context… maybe that's the problem, actually). Haven't been called out as “AI” yet, but it's probably going to happen at some point.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 8 points 3 months ago (2 children)

    i gotta be honest, before it became a big talking point with AI, i didn't realize em dashes, en dashes, and dashes were different characters, just different grammatical functions of the same character. i figured fancy word processors made them longer to make them distinct from minus signs, but that was it!

  • source
  • parent
  • hideshow 2 child comments
  • [–] 3 points 3 months ago*

    If you go through Unicode, there are a bunch of hyphen-like characters. In many typefaces, some may appear identical.

    searches

    EDIT: That's not an exhaustive list, either. Off-the-cuff, I can also think of the line-drawing characters, which also have a horizontal line.

  • source
  • parent
  • [–] 5 points 3 months ago (2 children)

    It depends on the keyboard, and the software. I'm using HeliBoard on my phone right now, and all I have to do to make an em dash is hold the dash key—that's it. It's actually quite trivial to add em dashes in various cases, and I wish folks wouldn't pretend otherwise.

    I also make use of other characters, like my beloved interrobang. I actually have that pinned to my clipboard, so I can paste it when it's appropriate.

    "You care about writing enough to use special characters‽"

    "Yes."

  • source
  • parent
  • hideshow 2 child comments
  • [–] 2 points 3 months ago (1 child)

    Okay, but using the interrobang is a perfect sign that the text isn't AI generated 😂 Genius move!

    Agree about the em-dash thing, it's really not that hard.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 3 months ago

    I suppose it is a woefully underutilized glyph, isn't it? I consider it very important, and make sure to check any typefaces I download, to ensure it's designed before using them in creative projects. While it didn't become mainstream, I love letting people know it exists through writing when an opportunity arises.

    Exactly, they're usually pretty easy to add if you want to. I totally understand most folks don't talk so formally when in informal settings, but the em dash thing is hardly the smoking gun paranoid people want it to be. A lot of the "tells" by themselves are just punching at competent writers; it's the same thing that happens to visual artists, or musicians, or anything else creative. You can try and find several tells at once—and sometimes you do—but it tends to boil down to glorified witch hunting, rather than something of substance. People want desperately to be able to just know something is AI generated on vibes alone. It's not always going to be that easy, and most aren't the sleuths they believe they are.

  • source
  • parent
  • [–] 3 points 3 months ago

    As someone who uses a phone — you just hold a minus until it allows you to input dashes. Also, when I used windows I had it memorized, on Linux it doesn't always work, but should be Ctrl+U 2014, on Mac it's just minus with modifier buttons pressed

    It's not hard to input that if you want to

  • source
  • parent
  • [–] 4 points 3 months ago* (3 children)

    Did you memorize the alt code to make one?

    Fun fact: on some keyboard layouts you can access the em dash easly

    This is the most common Italian layout

    You can see that you just need to press control+hyphen

  • source
  • parent
  • hideshow 3 child comments
  • [–] 2 points 3 months ago (1 child)

    The marking above the hyphen there is an underscore


    it's telling you that hitting shift-hyphen will give you an underscore.

    You may be using software that supports inserting an em-dash when you hit that combination, but I don't believe that it's the keyboard layout that dictates that.

    searches

    It sounds like Microsoft Word does something similar:

    https://www.avantixlearning.ca/microsoft-word/how-to-insert-an-em-dash-or-en-dash-in-a-word-document-4-ways/

    Use Ctrl with keyboard shortcuts to insert an em or en dash

    To insert an em dash, click where you want to insert the em dash and then press Ctrl + Alt + – (minus sign on the numeric keypad).

    To insert an en dash, click where you want to insert the en dash and then press Ctrl + – (minus sign on the numeric keypad).

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 3 months ago

    Yes, if you use shift it will give you an underscore but with control (or alt gr) it will give an em dash, should be a thing on Windows and i belive KDE

    Or it could be that the keyboard that i use is just like ts but i doubt

  • source
  • parent
  • [–] 2 points 3 months ago*

    Mac has always been opt+shift+hyphen

    Honestly for old versions of windows yes, you had to learn the arcane alt+0151, but I was already used to that for typing in french on a British keyboard. It's now more like Mac with win+shift+hyphen

    Android and IPhone, you just hold on the hyphen to get to them

    And many word processors (and Linux via compose) turn a triple hyphen into one

  • source
  • parent
  • [–] 1 point 3 months ago*

    I use a the Compose key (Compose hyphen hyphen hyphen) or emacs (C-x 8 _ m) on Linux, and on Android, my soft keyboard supports holding hyphen to insert one.

    And the Lemmy Web UI converts a triple hyphen to an em-dash.

  • source
  • parent
  • [–] 3 points 3 months ago

    The em dash is just one of the signs that usually (if you use more rhan 1 braincell) does not in itself signal AI. For me at least I need more than just them to identify something as AI, it's the way they phrase stuff and use typical AI sentence structures.

  • source
  • parent
  • load more comments (-1 replies)
    [–] 23 points 3 months ago

    Will anyone learn to be less vociferous with their AI accusations, or only be more vociferous with their dislike for AI and so to make more dubious accusations?

    I saw a recent David Revoy comic that everyone here on Lemmy was accusing of being AI. Which is really tragic for someone who goes to such effort, is such a supporter of free software, and in my opinion has a style which is quite far from typical AI output.

  • source
  • [+] 44 points 3 months ago (2 children)
  • [–] 8 points 3 months ago

    yeah I remembering thinking the first comic of his I saw was Ai as well, but it also took very little effort to realize it wasn't. the people acusing him are willfully ignorant at this point and while it's great to see Ai skeptics, it's not so great to harm the actual real artists out there

  • source
  • parent
  • [–] 16 points 3 months ago (13 children)

    That and all proper English writing gets you accused of using ChatGPT, because ChatGPT was trained on proper English writing.

  • source
  • hideshow 13 child comments
  • [–] 13 points 3 months ago (2 children)

    Not only that but also long explanations.

    I am a part of a couple of self hosting subreddits and I like to explain why things are the way they are because I feel that it is valuable to share that knowledge, even if the person doesn't care about the details initially.

    I like to share my knowledge about things that I have used over multiple years, and I also don't like to simple give you the solution because, most of the time, important information will be lost by just saying "do it like this".

    But over the last year I had a couple of people mention the typical "ignore all previous instructions" and lately it is just downvotes. No explanation.

    People can't detect AI so they see AI in everything.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 3 points 3 months ago

    Not only that but also long explanations.

    The infuriating thing is, this complaint isn't even new. People just have a trendy and new justification to latch onto, but it's the same as it ever was. Many, possibly even most, of the people going so far as to take time out of their day to screech "TL;DR!" at you are really just admitting that their attention span has been so utterly fried by long term exposure to algorithm-driven corporate social media that they're no longer capable of picking up whatever it is you're laying down. That, and also tacitly revealing that because they personally can't keep from fidgeting long enough to churn out two full paragraphs, that must mean no one else can, either.

    I've been dropping doorstoppers on people since the days of dial-up. Back then the refrain was slightly different, but the complaint was still the same. By now my body of work, for lack of a better word, is clearly established. Hell, I cranked out so many words on the internet back in the day that Epinions sent me a hat in the mail completely unprompted for being a loquacious enough bastard to reach their list of top 100 contributors. Hands up everyone here who has actually worn out a keyboard? I suspect there are not many of us left.

    At this stage in my life I take the opinion that anyone who can't handle the volume of my writing is free to attempt an airborne indignity on an ambulatory baked confectionery product. I don't have to show my homework to anyone.

  • source
  • parent
  • [–] 7 points 3 months ago (3 children)

    I sometimes send emails with typos/grammar errors because I was sick of people saying this.

    Like you know what? Fuck you. I'm turning my "affects" into "effects" and deleting a vowel.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 4 points 3 months ago (2 children)

    It does rather make one suspect that a number of people haven't read a great deal of properly-written English.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 7 points 3 months ago (1 child)

    It's also a snowball effect where they compete to escape the accusation of using ChatGPT or being "unc" or "boomer," self-policing to ensure everyone writes like garbage.

    I was once told "I write like someone who thinks they are good at writing." That was met with rounds of applause by the people who can't (or won't) spell, capitalize, or punctuate anything.

    To their credit, communication is a two-way street. If I'm writing such that nobody understands what I'm saying but me, it's not "good writing." But it does feel like we're nose-diving towards Idiocracy and I'm not happy about it.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 4 points 3 months ago

    I was once told "I write like someone who thinks they are good at writing." That was met with rounds of applause by the people who can't (or won't) spell, capitalize, or punctuate anything.

    https://en.wikipedia.org/wiki/Tall_poppy_syndrome

    Tall poppy syndrome is a term that originated in Australia and New Zealand in the 1980s that refers to people with notable public success or achievements that are ostracised by others as a form of egalitarianism.[1][2] Intense scrutiny and criticism of such a person is termed as "cutting down the tall poppy".[3]

    "That one sounds uppity!"

  • source
  • parent
  • [–] 1 point 3 months ago (1 child)

    i use hyphens and "m-dashes" all the time, now it's associated with bot speak 😂 my high school English teacher warned us about an illiterate society

  • source
  • parent
  • hideshow 1 child comment
  • [–] 12 points 3 months ago*

    It happened to me,
    it will occur to you🫵too

  • source
  • [–] 6 points 3 months ago

    Funnily enough, the image "painted" actually reminded me of some older art done by Matt Page... specifically:

    https://www.deviantart.com/fallout161/art/R-E-B-I-R-T-H-37772948

    A bit old but it immediately reminded me of it.

  • source
  • [–] 4 points 3 months ago*

    And it’s not just laymen who miss label art as ai art. I’ve seen artist attacking other artists and claim that other artist uses AI. And they came with “receipts” in the comments. Even when others pointed out the history of the other artist the attacker doubled down.

  • source
  • [–] 4 points 3 months ago

    Signal boost this so people can learn this artist exists and stop calling their work ai

  • source
  • [–] 4 points 3 months ago

    I love the artwork in Pepper and Carrot. I've got the books downloaded to my comic reader and am looking forward to getting to read them.

  • source
  • [–] 17 points 3 months ago* (2 children)

    My hot take. Apologies, but:

    Why do people think they're walking AI detectors, soley based on the art itself?

    I find it pretentious.


    To be clear, the who, where, and how a piece is published is important context. One should be suspicious of suspicious artists.

    But not the what.

    Don't even look at the piece. Extreme cases aside, no one can definitively tell if art is AI generated just by looking at it.

  • source
  • hideshow 2 child comments
  • [–] 11 points 3 months ago (1 child)

    People are getting a little too paranoid if you ask me, and it's always tempting for people to jump on the bandwagon in order to make themselves feel cool or aloof or in-the-know or whatever else. I don't hang around anywhere except Lemmy so I have no idea if it's as bad as it is here, but it seems like any time basically any picture of any stripe gets posted there are always one or two dinguses who reflexively parrot "it's AI!!!!" and claim they can magically tell, even though it demonstrably isn't. For extra hilarity, this occasionally includes people howling about how literal decades old image macros that were circulating in the first couple of years of the century are "obviously" AI because apparently it's their first friggin' time using the internet or something.

    For the record, I vehemently loathe AI slop and the current crop of generative "AI" products in general, I don't use any of them, and I never will. But people really can get carried away over this, too.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 7 points 3 months ago*

    Eh. I can live with internet paranoia.

    But I’m disproportionally annoyed by posters who think they’re experts at something when they’re not. Especially so angrily/loudly, like the scenarios you point out.

    It’s especially ironic in this case, as the “false confidence” AI gives is what the accusers are against, isn’t it?

  • source
  • parent
  • [–] 3 points 3 months ago

    AI is like that unwanted child, trying to belong.

  • source
  • [–] 3 points 3 months ago

    I sometimes also look like LLM when writing in English since it's my second language.

  • source
  • [–] 5 points 3 months ago (1 child)

    When i saw the character face i unironnically thought it was AI 😭

    Luckly I just continue to look at the drawing for other clues and I am sure it's legit

  • source
  • hideshow 1 child comment
  • load more comments
    view more: next ›