Is AI Ruining Our Perception of History?

[@Dejmo]

I wish Google had an option to filter out generated material.

[@poe8910]

that would be great but i don’t think it will happen. they are pushing this “AI isn’t that evil” agenda so they can make more money, especially given they have their own AI model (which is horribly biased)

[@dragonflower17]

if you type -ai in the search bar it removes results that feature it. It's not perfect, but it has drastically cut down on it for me. If ai isn't listed anywhere on the page it's found on though.... you're still gonna have to look to make sure. Another tip I've seen is to use the date filter and only search for results before 2020.

[@JutaLovelace]

Right now the general rule is to type -prompt in your search.

[@aliciabell6688]

Loving your boolean logic. ❤

[@SysterYster]

Problem is, a lot of people don't write/tell it is, so how should google know?

[@skeinofadifferentcolor2090]

As a crocheter it's becoming increasingly more concerning seeing how many AI images of crochet items have been infecting the community. So many people don't realize that there is no pattern to these projects because it's computer generated, and it's so disheartening as a creator being taken advantage of by a computer program that misleads people into thinking that such projects are achievable.

[@streetshoes]

I've seen this for sewing too

[@shoszannah]

OMG I just needed to unsubscribe from serval fb crocheting groups because of that😮 Everyone "older" started to share and comment on AI generated pictures of crochet impossible to make. I tried to educate them but it's a lost cause and it's SO SAD

[@Narangarath]

@@shoszannah I feel like those people's children and/or other younger family members have completely failed them. I mean, I get it, it can be hard to pick up all this new stuff being developed without having someone help or stumbling onto a particularly good source of information with side-by-side comparisons and extremely detailed explanations. And it doesn't help that many (not saying you by any means) people who point out things being AI generated do so in a supremely condescending way, because who's going to be comfortable asking for help with something that makes them feel, and be talked to about, like an idiot?

[@eldrago19]

Sorry, did you just say it's impossible for a computer to generate a pattern? If you really believe that, I'm not sure where to start.

[@songindarkness]

@@eldrago19The point is that there are lots of AI images of impossible to make sewing or crochet items out there for which a pattern does not exist and this is misleading people.

[@brodig]

I'm almost finished my master's of library studies, and currently work at a public library. People are always asking why libraries still exist when you can just google something, but stuff like this is a major reason why. We're trained to get you accurate, relevant information from reliable sources while google is designed to get you a lot of information from a lot of sources. Even I was surprised just how much of my program is focused on negotiating AI in information spaces. Support your local libraries! We're so happy to help.

[@Pollicina_db]

Thank you for your work, keeping our history safe is really important.

[@vickywitton1008]

❤

[@view-camera-queen]

Totally! Libraries are also community and learning spaces. So much depth to what they offer and they deserve so much more money and support

[@mariebray9831]

Yay for public libraries.

[@Lonovavir]

As a WW 2 reenactor I can confirm how essential books are. Not only do they help LARPing nerds like me know what we're talking about, but they also have photos we base our kits on. With an internet image you can never tell if it's a real photo a couple in the 1920s or AI. With a book you're more certain it's real.

[@Sunshine-my1vd]

The past was erased, the erasure was forgotten, and the lie became the truth. George Orwell, 1984.

[@Silent_Depths]

An ordinary day like any other. Happened time and time again, today's no different.

[@mroxannevh]

And if the technology has been around as long as her research says it has then they had decades to regulate it. But they did not and it's been on the down low

[@songindarkness]

⁠@@mroxannevhAI experts have been sounding the alarm for years. But the big companies have fired many of them and/or and buried their concerns in NDAs. Or they like to divert people to the fear of sentient AI which deflects them from worrying about the real concerns of the generative AI now. There were some good articles in Wired magazine about this issue recently.

[@songindarkness]

It’s so scary. We are definitely in a disinformation age. We are in the early days of a cyberpunk dystopia.

[@BlackSeranna]

We’re seeing it happen right now.

[@alicias.8482]

I wore a Victorian outfit to school to read A Christmas Carol to my students, and a kid asked me if I was dressed like Mary. Like from Nazareth.

[@maleineperle1770]

😂😂

[@RogueVideoRaven]

Well… were you?

[@veracrow842]

Victorian core Maria

[@sharbean]

Ignorance is rampant.

[@charlottec5802]

Something similar happened to me when I was dressed in a Regency outfit. OK, not everyone can easily place an outfit in the correct time period, but seriously?! 1,800 years out?!

[@rudetuesday]

I worked at a big-chain bookstore before the current wave of A.I. took hold. So many customers didn't have a keen grasp on when things happened in history, and this affected what they asked for: books with photos--not paintings--of Queen Elizabeth I and audiobooks of The Bible featuring Jesus's voice. These people weren't trolling.

[@The_Super_NOVA]

LOL audio Bible narrated by Jesus 😂

[@haltersweb]

I love the many translations of the Bible and love viewing them inter-linearly with the original Hebrew, Greek, and Aramaic texts that they were translated from. One day someone mentioned the use of the King James translation: “Aah, the King James Version. Finally people are going back to the original source.” 🤦‍♀️

[@alisonmercer5946]

Bible narrated by jesus no way they wernt joking

[@rudetuesday]

@@alisonmercer5946 I was in that job a long time, and come from a very religious community. I've met a lot of different kinds of people. They weren't joking, although I really wanted them to be.

[@GataMiau44]

Wow. Just…wow.

[@erinoflaherty8756]

"It does not crave to awaken feelings in people, unlike artists." You hit the nail on the head right there.

[@OctopusOwl]

Such a perfect phrase

[@TheManinBlack9054]

If it doesnt, but it still manages to then what?

[@tamarakonczal6350]

Fabulous.

[@ember9361]

@@TheManinBlack9054 because it doesn't manage to. That's literally all pareidolia. Nothing else.

[@gunstarheroine5173]

This is the best way I've ever heard it put

[@arianalheureux7290]

I think you touched on another reason why people use AI, and it’s the yassification problem. It’s hard to source an actual historical image that’s relevant to themes sometimes, for sure, but our concept of beauty is now so stilted that for the purpose of fashion or historical fiction, a super yassified, super chic (by modern standards) image is going to be more attractive to modern sensibilities than an actual historical image. It doesn’t feel like people even know what people look like anymore.

[@lucie4185]

An author I like has started using AI images for her covers because it's cheaper and I just feel so icky seeing them.

[@arianalheureux7290]

⁠@@lucie4185That’s so sad :(

[@user-to4fo2xb8v]

It’s like what they do in movies when they keep their 2020’s eyebrows

[@lucie4185]

​@@user-to4fo2xb8v Tans, teeth and brows.

[@arianalheureux7290]

@@user-to4fo2xb8v And barrel curls and balayage. lol

[@dariuslateralus7128]

This means that even small museums and archives will become more of an authority again, as they have old physical documentation and carefully preserved information.

[@SamAronow]

Only for the few who go looking for it.

[@KeiPalace]

we can only hope, this also merges with the drive against intellectualism and the dismantling of our education system in the US, people who are kept ignorant are easier to manipulate, at least the younger generations are more savvy.

[@patrickhector]

​@@KeiPalaceironic, I get the impression from a lot of anti-ai people that they hate the technology, which is inherently anti-intellectual

[@sarahbeckwith3087]

​@patrickhector I hate the technology because it makes me feel like I've been conned. Like 'it' has insulted my intelligence

[@dariuslateralus7128]

@sarahbeckwith3087 it's still just a tool. It's the creativity of con artists, which makes it so dangerous. That is exactly the reason why we need well working legislation around the globe. What museums like mine can offer is a collection of originals, that, if curated correctly, can be a reliable source of information.

[@missrobinhoodie]

"it's tempting to get exactly what you want" is so fitting for life in general and social media in particular! As Oscar Wilde said: “There are two great tragedies in life. The first is not getting what you want. The second is getting it." Life is suprising. History is suprising. And suprises are usually not exactly what we want, and they might not even be pleasant. But that's the beautiful secret of it all!

[@E_FoxSnowspirit]

Yes!!!!!

[@eva1777]

We really went deep into the misinformation era. Cannonballed right in.

[@Bionickpunk]

We just started to be informed directly from people on ground level as oppose to traditional news centers like TV or newspapers. It won't take long until you can't trust even news coming from "the people".

[@KeilaBevins]

Delulu is the solulu fr

[@nicholauscrawford7903]

Most of the world has been deceived by Satan about most things of eternal significance for millennia already.

[@Bionickpunk]

@@nicholauscrawford7903 Oh please, spread your religious sect nonsense somewhere else.

[@The_Super_NOVA]

Nah, cannonballs are epic. We silly salmoned into that tar pit with foolish abandon 🐟

[@Sophie_Cleverly]

I'm a children's author and I've already received a couple of projects from kids where they obviously asked AI to tell them about me, and it's told them that I actually have 3 kids or that I have a dog or that I live in a different city or whatever. It's really bizarre and I wonder what people will think about me in future with all this misinformation. Obviously this is a super low stakes thing but it seems like such a slippery slope to awfulness.

[@songindarkness]

This is awful and really scary. Imagine if the AI had told them wrong information that actually put you in a bad light or worse? Careers could casually be destroyed by a kids project.

[@E_FoxSnowspirit]

A very slippery slope indeed

[@KyoShinda457]

That's true. Well, hope things go well with your 3 new kids and pet dog.

[@giovanaconod8129]

That is not low stakes, misinformation about your personal life is being spread and could be used against you to harm you, you matter! I'm very sorry this has happened, and I wish you lots of success

[@lainightwalker5495]

i have recently seen alot of history channels useing ai-generated images when talking about historical persons and it dosent clearly show its ai generated. it should be.

[@matheussanthiago9685]

Every KZbin video should come with the disclaimer that it contains AI generated imagery/ content That way I wouldn't even bother to click any of them

[@lainightwalker5495]

@@matheussanthiago9685yes i think so too

[@meikewassmuth4801]

I've started to block more and more channels, where I strongly suspect AI "art" to be used without disclaimer.

[@lfgifu296]

some do, though- still hope in the world😭😭

[@allana1997]

I’m not opposed to that use of it but it just atleast have like a caption that says AI GENERATED IMAGE or it should be verbally mentioned

[@LarnieRadek]

REALLY loved your point about A.I. not understanding posture! As it only understands everything face value and from images alone, its pictures can’t reflect how structure, undergarments, and even etiquette is affecting people’s poses and movements. It only sees people and clothes but doesn’t understand how they are interacting with each other and their environment.

[@Realma]

AI and the algorithm are the main reasons why I'm spending less and less time on the internet. In fact, I feel like the internet itself is turning into this big bin of AI-generated material that is there to feed the more and more insatiable algorithm. I want to see real art and creations made by inspired minds and hearts. Those imagined and crafted by fellow creators, artists and musicians with the goal to touch our souls. Not soulless products that are solely there to chase the views and in the process taking our attention and perception hostage...

[@Smethells2023]

Check out the Dead Internet Theory if you haven’t already. Roughly 60%+ of online traffic is just bots. It’s not going to get better in the future, exactly because of the algorithm and the endless need for content (even if it’s low-quality, inaccurate, and pointless) plus the advancements with AI. Fun times!🙃

[@Realma]

@lls2023 Oh wow 😅I guess the internet is slowly entering its 'live long enough to see yourself become a villain' era...

[@matheussanthiago9685]

Frfr

[@Natalie-101]

That sounds incredible! I think the exact same thing yet the algorithm has its hand so strong around my neck that I can't peel myself myself away, sometimes for 10+ hours a day. I've been trying to actively spend less time on the internet but I often fail, especially when things like school and work require me to always be on my computer and often on the internet itself. The pull to the endless scroll is just too strong. Any advice is welcome!

[@kittykittybangbang9367]

​​@@RealmaThe internet was supposed to be digital anarchy, but then corporations ruin that. And let's be real here, the same thing is probably going to happen to AI and VR in the future so you better enjoy it right now because corporations are definitely going to ruin those two things.

[@datpotat3945]

15:21 this is the thing that gets my goat so much. People argue that the ai image generation process is the same as any artist's creative process,.like, excuse me? The day that AI comes up with its own style, technique, approach, and philosophy, and an original artwork to represent it *unprompted*, is the day I'd agree.

[@FunSizeSpamberguesa]

I had a lot of fun recently messing around on ImageCreator while sick (mostly because I like seeing all the ways it screws up, like giving people three feet and, in one case, a guy with bare disembodied legs growing out of his groin), and I noticed that no matter what physical descriptions I put in for women (age, build, skin tone, facial features), I got the same pale, plastic-faced/giant plastic tits 20-year-old Instagram chick with fake fat porn star lips every damn time. It sometimes does *slightly* better with men, but it's still tough to get anything other than 'bland GQ model.' In a way, though, I think that might be a good thing, because it's a dead giveaway you're looking an an AI image.

[@kaiyodei]

it gets worse when people weaponise disablity. "I'm disabled this is the only way I can express myself, it's a form of creativity" or "i bet you don't even make your own paintbrushes and paint, stop hating on AI"

[@gargoylekitty]

@@kaiyodei Knew a guy waaay back on /ic/ that lost use of his dominant hand and tracked his progress on the board using his weaker hand and he got just as good then better. There are artists with no hands who use their mouths or feet among a number of alternatives. It’s not disability but lack of willingness to put in the effort to learn commingled with desire for instant gratification and praise.

[@thirdeye4654]

What you're asking from a model is agency. I mean we're not far off. "Unprompted" is a weird ask though, because this correlates to an artist without a brain or body at the moment. ;) I think some people overestimate what human artists actually do... Bringing their brain dreams to life.

[@nimue325]

@@gargoylekittyThat’s important. The artist Chuck Close went from a hyperrealistic style in the 70s to a portraiture which took pixelation and color theory to the next step to adjust to his physical disability, and that is a single layer of meaning on his work. AI may be able to create an image but it cannot create a “why,” which is necessary for something to be art, in my opinion. I also see a steep rise in use of the misogynistic term “suffragette” these days, instead of the term used by suffrage fighters themselves, suffragists, which reminds me that expertise rather than mere usage is necessary and being lost these days due to the spread of both AI and social media (and thank to channels like this that do help with better education!). At some point, if AI generates mistaken statements of images often enough, will it begin feeding on its own misinformation more heavily than on real history?

[@johanna2460]

in addition to all of your excellent points, one of my biggest criticisms of ai (and I am by no means an expert) is the way it is trained, which is mainly by consuming actual art and writing to then cheaply reproduce it and ripping off the original creators who oftentimes spent hours and hours on their art and years honing their craft only to have it stolen by AI

[@mihaleben6051]

Solution; stop spending so much hours on art.

[@innovationsanonymous8841]

Solution; make something new. AI is for reproducing what's been done.

[@mihaleben6051]

@@innovationsanonymous8841 like uh... a book about living blankets.

[@E_FoxSnowspirit]

@johanna2460 thank you!!!

[@bossyboots5000]

Same. We're not seeing the miracle of AI, we're seeing copyright infringement, plagiarism, and theft.

[@pascat3104]

National historical museum in my country completed an unfinished historical painting from the 1800s with midjourney... The museum argued that it was fine because "the artist would have used the technology anyway if he had been alive today." I almost cried when I saw the announcement.

[@lunayen]

What? I can't believe that museums are actively working with AI to doctor the truth. Unfinished pieces are unfinished for a reason; a museum who won't let their visitors imagine why a piece remains unfinished is a terrible museum.

[@wayIess]

On one hand, I think any creator would love to snap their fingers and have technology auto-magically complete a project. But there's so much value in the process. Time, effort, blood, sweat and tears. There's definitely a problem with the museum ignoring all that.

[@durd320]

​@@wayIess tbf if one ever actually picks up a pen or a brush to paint or draw and really tries to understand the fundamental, they would instantly realize why artists dont generally consider image generating ai models to be in the realm of making art, the process of creating forms and things that can be perceived from just stuffs on a blank canvas is fucking magical man, pressing a button/typing a prompt is not the same,....it's like searching on google image but more accurate, a museum failing to understand this is disappointing lol.

[@magda5352]

I think the museum has bad reasoning as obviously that is not something that can be proved and they should base their assertions on evidence. However I do not see what's damaging about creating a projection of how the piece may have been finished. No one going to the museum will mistake it for the real art as one is physical and the other is on a computer. The techniques the artist used are not lost because of a digital experiment. One of my favourite installations is a 3d printed greek statue but coloured in, they way we know theirs were. This was done with a computer and does not take the same craftsmanship that painting a marble statue takes. However it is really interesting and does add to the historical narrative. obviously I'm unaware of the particular work Ur talking about, however it seems unlikely that they would now be present the ai art as "this is the art", but rather it's a creative experiment (done well or not) which imo does not harm the original art work.

[@durd320]

@@magda5352 the thing is, compiling a bunch of images and spit out something that seems passable or visually appealing isn't representative of any original visions, the example you made with the greek statues was done by people with actual research and using that kind of research to recreate the original vision, these "ai" models are not actually ai, they do not have access to any linguistic or intellectual datas, if you tell them to generate something like a classic painting from the victorian era, they wont be able to understand the context of which these paintings existed in, theyre only basing the final image on keywords and that is not effective at showing what these paintings couldve looked like. And fyi, retouchers exist, people have been doing art restoration for DECADES now, and they understand what they're doing, unlike relying on some random probability

[@VoltasP]

I was playing around with one of those ai tools and seeing if it could generate accurate images of Edwardian clothes-- Because I was asking this same question-- How misleading would AI be to aspiring cosutumers who would rather let AI do the work? Shirtwasists and stuff, you know? Absolutely nothing too difficult. And there was NO WAY to stop it from putting BRIGHT RED lipstick on every single mature feminine-coded individual in a full-color generation (as opposed to generating B&W photo type images). Even when I put "lipstick" and "red lipstick" in the negative prompts is was like "Sorry, but bright red lipstick is non-negotiable". It wouldn't even do pink. Just bright red. And it was such a weird minor bug, but still a puzzling one, and I know that AI tech bros give, like, negative f*cks about fixing 1910s historical makeup accuracy, but I can't help but wonder if this is one of those butterfly effect scenarios where every movie and TV show from now on set in the Edwardian period is going to have 1950s lip shades on all the women.

[@wickedthing6068]

I feel it’s definitely because of the movies/tv shows

[@Nassifeh]

@@wickedthing6068 It's, um, worse than that. None of these tools has been trained on movies and TV. There are occasional stills from movies and TV in their training data, and some show up more than others depending on how much it was promoted? But people overestimate the influence of that, I think. Instead, take a second and just think about, if you weren't basing your training on any video, if you were just basing it on images scraped at random off the internet, imagine what a substantial portion of those images are, that would over-train the model on generating women with bright red lipstick. In recent models they've done a lot to try to get them to stop generating that kind of thing outright, but it required actual effort because even if you weed out the most explicit stuff, the model is still overwhelmingly trained on pictures of cis women that were taken primarily for the enjoyment of cis heterosexual men in a very particular way.

[@thing_under_the_stairs]

@@NassifehHoly crap, I never thought of that! Thanks for pointing out the horrifying reason that AI generated women never look real!

[@merrimcarthur7198]

I'm 70. I am constantly pointing out AI nonsense to people much younger than me, who then disparage me for harshing their buzz (as it were). Indeed, ALL my older friends (anyone over 55) point out AI b.s., again, to younger people. Do NOT assume that all us oldsters are ignorant, calcified naïfs who are easily duped. Equally do not assume that all the young folks know whereof they speak regarding tech. Of course, perhaps I'm a bit of an outlier in that most of my friends ARE in tech. I've worked in tech (low level IT)...and entertainment. So, it's easy for me to spot fakes. What amazes me is that young people are always assuming I don't know what I'm talking about simply because of my silver hair. Regarding AI in "film", that is a potentially terrible thing. Same goes for AI in research, medicine, design, etc. You are absolutely correct in calling it soulless. Personally, I wish current AI did not exist. The tool is rather neutral, but how it can be used is not. Good video.

[@ericahinman8416]

Thank you for pointing this out. I'm Gen X (specifically xennial) and I think that gives me a unique perspective from having an analog to digital childhood. The younger generations are painfully vulnerable to misinformation but do not have the touchstone of pre digital experience to help sift through the BS.

[@Joy2Life333]

As a 60 year old woman with a background in libraries and tech, I thank you for saying this so eloquently. I have helped many a young people in research. I know how to dig into source materials.

[@atinycrow]

She never said ALL OLD PEOPLE. It's a data-backed correlation between increase in age and likelihood to fall for fake news on Facebook.

[@cocopersiflage4705]

@@atinycrowcould you source that? Specifically related to middle-age people (40s/50s)? It’s ironic that we’re having this conversation since part of this video is about not perpetuating myths

[@sinxenon3181]

I'm young and I can compare your generation with mine ~30 and younger generations. And I can say without any doubt that your generation is much smarter and well-educated than the young ones. I don't know what to expect from the future with these young and naive people around. I hope, you guys will live long lives and help us to not f*ck up.

[@katk925]

I keep seeing people share "vintage" pictures they think are real and I'm like yoooo in the 60s people didn't stand in the background with their hands up like they are taking a selfie

[@kaiyodei]

I think I saw one of Yasuke with...some woman.

[@ames522]

"stuck in a shrinking circle of reheated ideas" -- nailed it.

[@isthatachicken]

It's warping _everything_ . The amount of fake anime stills on Pinterest and just randomly generated pictures of celebrities on Tumblr? It's a bit sad to see people believe those images (and then be sad upon realising it's Ai generated) because it means many can't tell the difference that well yet? Edit: Now the warping of _literal_ history 😭😭 please I don't want to second guess everything??

[@AnEmu404]

AI is getting better, but people struggle to tell the difference either way. I saw a yt short pop up that was OBVIOUSLY AI slop - it was “history” that had a robot voice, nonsense script and AI generated fake historical photos. Thousands of people in the comments were talking about it and debating as if the video was legit, not one person realising it was AI. Terrifying

[@rhubarbara2877]

It's giving kind of 1984 vibes. Just imagine what the Ministry of truth could have done with AI technology

[@kittykittybangbang9367]

​​@@rhubarbara2877I'm really scared of what governments and corporations are going to do with AI warping history. They can easily use it for propaganda.

[@ember9361]

@@AnEmu404 people struggle with manual image editing (think super slim unrealistic waistlines, "photoshopped" with paint to exaggerate the figure that people today believe its real)

[@janebeckman3431]

@@AnEmu404 I've run into several of these with anachronistic details in the AI art.

[@kb-ny3ln]

Pinterest has become full of Ai images, and I'm pretty good at picking our real from AI, but the ones that really trip me up are the fashion and interior design images. An artist I follow (I can't remember who) said they were struggling to find reference images for certain animals because even science articles had started to use AI images. It's very annoying trying to find examples of stuff your looking for without using critical thinking skills to determine if it's AI all the time...

[@пекельніборошна-т1в]

Researching art is a nightmare now. You type the name of the artist in Google search and the search is full of AI promts like 'Klimt style painting'. Some are rediculous and onviously AI, but some make me stop and think (I'm not an art historian, but studied european art for over a decade as a hobby). This is just devastating.

[@snooksmcdermott]

Do a Google search for art history LibGuides

[@Pantheragem]

About 12 years ago, some news broke about Wikipedia and editing, etc. At the same time a lot of people were getting rid of their "outdated" encyclopedia sets in favor of finding info on the internet. I, along with others, urged to hang onto those encyclopedias because one day they'd be more accurate than what you were finding online, even if outdated. Wasn't expecting ai to screw history up even further I guess.

[@sarahmcmullen5938]

AI being mistaken for real things is very concerning. I question everything anyway... but I've learned that since digging deeper into history. Finding the context/frame of reference/goal of discussing or exposing whatever it is I'm studying often helps me question and dig deeper. I just am afraid of what happens when iterations upon iterations of AI continue to change the story so we really can't get to the bottom of what is real and what isn't.

[@SplatterInker]

When we lose the art of citing sources...

[@lucianaromulus1408]

​@SplatterInker half the sources are crap too lol. Getting my masters degree really helped it sink in for me. Even supposed experts have crap data, let alone this AI 😂

[@hhoi8225]

This is why it matters to own real books. Even boring books, and bad books, and books that one doesn't like or agree with. They can't be edited in post. 🤷‍♀️

[@lucianaromulus1408]

@hhoi8225 speaking of AI I see MULTIPLE comments have been deleted. This isn't even a controversial topic...but I think it highlights the possible evil behind AI given who programs it.

[@user-gu9yq5sj7c]

@@hhoi8225 It's easy to say to say that but owning a lot of books or having space for all those books, and being able to travel to libraries frequently or many museums, etc can be expensive and excludes poor people. It's also time consuming to look through books. I saw comments saying they appreciated yt and its audio because they're blind. Audiobooks are costly too.

[@rdreher7380]

Two things come to mind for me. First is a comment my model railroading Dad often makes. Often, when we go to train conventions and see models of buildings on the market, he will comment how something strike him as a "model, based on a model, based on a model..." rather than a model based on a historical reference. Like you are for clothes he is for buildings, signs, all the sort of details that go into a making a model railroad meant to depict a certain place or time, and his point is you can kind of feel it when something is more of a model railroadism, ie an interpretation of an interpretation of an interpretation of history, by the nuances of how fanciful or romanticized it is. The other thing that comes to mind is when I and my friend were having some fun testing prompts with ChatGPT. Both of us had lived in Japan for many years, and were joking about some funny awkward translations we had seen, and wanted to find out if ChatGPT could understand how to replicate what gives "Japanglish" it's particular quirky charm. I tried asking ChatGPT to give "funny, ungrammatical" sentences. The results were really disappointing. It seems to have latched on to an idea of "grammar" based more on what annoying fuddyduddies call "grammar," ie proscriptive rules of what "good English" is and isn't. Worse yet, the "ungrammatical" example sentence felt like stuff that would just be normal African American or Appalachian English, in other words simply non-standard English, but not really weird or funny mistakes. Thus, ChatGPT came across as basically classist/racist. When I tried instead to say "sentences with syntax errors" it made some really funny examples, ones that actually broke the structure of English and made for a good laugh. What I took away from this experiment, is not that ChatGPT failed to understand what "ungrammatical" means, but rather that it had learned, for better or worse, what "ungrammatical" means from us - the flawed human culture with a long history of classism and racism embedded into how we talk about "correct" language. The technical term "syntax error" is more specific to linguistics, and doesn't have the same baggage as "ungrammatical," so it got the result I meant. ChatGPT is just a mirror of humanity, it is learning from us, reflect back to us exactly what we are teaching it. It is our child. I see these two ideas connecting right into what you are talking about with AI generated art. Those algorithms are learning from humans, who learned from humans, who learned from humans, who learned from humans, maybe going back a long time with nobody along the way resetting their perceptions with primary sources. Just like with ChatGPT, those machine learning algorithms are being fed indiscriminately from all the available content of our modern global culture, and given us a reflection of it. I am at my heart both an artist and a teacher, and see these two endeavors as deeply connected. Right now, mass data farming machine learning algorithms, by way of acting as a horrid counterexample, are showing us what the value of a real artist, a real teacher, a real inspiration is. And until whatever apocalyptical future awaits us come to be, I'm so glad to have the chance to learn from amazing and creative teachers like you, who don't just make models based on models based on models again. Keep up the great work!

[@liamannegarner8083]

I can't help but remember the story of the AI that was trained to write Yelp reviews, that the researchers realized was giving Mexican restaurants lower scores - and then realized, since it was trained on English-language Facebook, the AI had assumed "Mexican" was a slur. AI programmed by humans will never be better than the humans.

[@wayIess]

Beautifully well said.

[@JadestonePony]

I think you figured it out very well. AI is in fact, a mirror. I'm a Data Analyst/Data Scientist, so I happen to know a thing or two about how AIs are created. An AI is literally only as good as the data it's trained on. "Garbage in; garbage out" is a phrase in the computer science world that's been around for decades, but it does seem like it's being forgotten over time. Regardless of how intricately developed an AI's algorithms are, if you feed it garbage data, it will give you garbage results. ChatGPT and all the other AIs out there are only reflecting our own misconceptions and false ideas back at us.

[@AnnekeOosterink]

This is the same issue as the program written to sort CVs into interview and reject piles. It rejected all women, because the programmers hadn't thought through what kind of material they teach. This happens time and time again. And often it isn't corrected, or barely corrected, and only after it already caused harm. Also it grinds my gears that we call what is basically a very fancy sorting program AI.

[@snooksmcdermott]

As a computer-adept Gen X rare books librarian, I frequently have been astounded by how little our students know about history. I get where you're coming from, but lemme tell you, many US kids are NOT getting history in school. Just an example - I have had to explain BCE vs. CE more than once, something I learned in elementary school (as BC and AD, back in the dark ages of the 1970s). So, I love ya, but there are plenty of us old folks who are not being fooled by AI, and plenty of young people who are.

[@slavetobloodcomic]

Seeing exactly the same thing. I don't think being fooled by AI content is really the preserve of older generations, but merely of those less savvy, regardless of age.

[@inkasaraswati7625]

I worry gen Z and younger would have difficulty in recognising fake images with how ubiquitous filters, and now AI images are. Millenials grew up recognising fake vs real images from all the terrible effects we saw from childhood, and we sort of grow up seeing better and better effects but we are still able to recognise telltale signs. I worry younger generations do not have the same senses because fake images are increasingly lifelike, or probably do not care if they're fake because they've grown up seeing fake images all the time.

[@diannemh2250]

Fellow Gen X here also feeling slightly miffed by the implication we are neither tech savvy nor able to question things!!! Understanding history, primary sources, questioning sources ( ie. is that actual peer reviewed data or just 1 person’s opinion?), making comparisons, questioning “ but why?” - I was taught these skills at high school & Uni; but I found when my gen Z kids “ digital natives” - went to school they were not necessarily taught how to do research online, how to question things. Their teachers assumed that because they are online all the time, therefore they somehow automatically know how to navigate this strange new world !

[@kirahowes9670]

Thank you; I came here to say basically this. Even the article referenced in the video specifies that Holocaust denialism in the US is more common among younger folks, according to recent surveys. Failure to think critically is a problem across the generations, even if it manifests itself differently among different age groups.

[@bluedragontoybash2463]

as a gen X... when I tell my teenage kids about european history. I have to constantly point out on map which is England, and which is continental Europe !

[@AlexisRohl]

My English class is actually studying the effects of AI on language and society’s perception of reality based on biased content and the points you made are so true and I completely agree with what you said about AI

[@deltaflute03]

I once went to a New Year's Eve party in an outfit that I made from a 1960s pattern. "Oh, how very 1920s of you!" the host said upon learning that I make clothing using vintage patterns. The 1960s silhouette looks nothing like the 1920s. I wasn't even wearing anything with a dropped waist. It was a sheath dress. I had to burst his bubble and correct him. Lay people, who didn't grow up in those eras, have interesting notions of fashion history. That's for sure.

[@ankavoskuilen1725]

I am an art historian and I am very worried about this, for all the reasons you mentioned. One example: I noticed that on KZbin there are some channels who illustrate their historical story from, lets say the 17th century, randomly with pictures from the 17th century but also with pictures from the 18th and 19th century. If all this is mixed up - and this wasn't even IA yet - after a generation you can't even discover what is true because your initial take on history is already flawed. You don't have the right yard stick to measure new information.

[@lfgifu296]

I feel like, much like with Period dramas but worse, you can’t believe what AI shows you when you look up “1637 high class woman” on mid mid journey (idk if it’s written together or not😭) because you will be conned. That reel of AI “historically accurate clothing” IS TERRIBLE- it angered me how God awful it is💀 the third leg in the middle of that was the least jarring thing in the middle of all that bs, and more often than not people (like me) who pointed out the errors were shit down because “it’s not supposed to be accurate and it’s well made, took a lot of work, so stop being negative” but it WAS supposed to be accurate and the fact that people believe it is is horrible and will lead people to idiocy

[@matheussanthiago9685]

The fact that a browser extension that detects and hides AI results haven't been made yet It just boggles my mind

[@Historyofstitchery]

“Took a lot of work” girl, it factually took zero work. That’s why they did it.

[@faameexplains1192]

What people don't realize is that midjourney isn't smart, it's predictable. It doesn't know what things are, it knows what images with that tag tend to have in common. I remember early last year seeing like "what if the ancient Egyptian empire never fell" and it was High tech mixed with ancient Egypt. What you would expect if you typed "futuristic" and "ancient egypt". But it was being passed off like the AI ran a simulation of the past 2000 years and took screen shots

[@georgiagrace1014]

Yes exactly!! People don’t seem to understand that AI generated images are PURELY visual. There is no “intelligence.” That’s why it always gets hands so wrong lol, it doesn’t have any concept of what shape a hand is in three dimensions or how it moves. It just has a wild guess at what a hand kinda looks like

[@storageheater]

​@agrace1014 The problem with it is that when we've stopped laughing at the extra fingers we'll see that AI hasn't consistently been getting the hands wrong for ages and when they ARE wrong you can run it through Photoshop's content-aware AI editing tool, or use Midjourney's Area edit thing to keep regenerating the hand until it's correct (you can automate Midjourney's regenerations so it'll do a load while you're busy, then you just pick out your fave, maybe filter it or edit on a few real details.) It's fun to cherrypick crap AI and laugh at it, I won't deny that. But the real joke is on us, that actually that we thought finding it funny would stop it getting very, very convincing to a lot of people, and it is getting better at a terrifying rate. Also you can just put "in the style western animation" and you suddenly have something that overrides our sense of concern about errors in the same way we don't criticise "real" art in the way we criticise Bing's attempt to create photos One of the most interestingly unsurprising things about the AI process is that it was sold to us as a product that saved time, where you could create your concept art and then hire real artists to finish it off. But of course, what actually happens is the creatives were removed entirely, and we have people who don't look critically at work at the helm, so the results are often getting roasted on social media. But the reality is, the tools are already *incredibly* powerful, and they're being used by amateurs. We're laughing at the amateur results and telling ourselves it's not too late.

[@lolarogers2313]

I appreciate you mentioning what is sometimes overlooked in discussions of AI--that all of its content is taken from things created by humans, and it's using their creations to make a fortune without paying anybody for their work. I recently learned that a bunch of my translations were fed to AI software without permission from me or from the authors whose work I had translated. That means that the glorious, weird, inspired ideas of the authors and the sometimes inspired ways I found to express them in another language have been sucked up by a machine to be recycled into soulless content for someone else's profit.

[@phoebexxlouise]

I would honestly watch a two hour video from you going through all the fashion eras of history creating a visual timeline as a reference guide for anybody in film and television to use

[@PabloEmanuel96]

But also, film and television have massive budgets so they could hire experts, or even Carolina, instead of her working for free on it

[@jasminv8653]

She actually has historical silhouette timeline videos already made, just go and find them

[@vf1923]

Such things exist quite readily, in the form of books (etc.) that focus on fashion history. But as Karolina herself points out, clothing is both local and personal. There's no such thing as a single timeline. Any costumer-maker worth their salt should know how to research appropriate fashion.

[@brendaokuda2158]

As a 60 yr old grandmother, I think the thing that irritates me the most about AI is how it blurs the line between real & fake. We can no longer trust our eyes or ears to know the truth. This highly disturbs me. To me it's akin to really enjoying a hamburger & then someone telling me I just ate artificially made meat. There needs to be a way for us to differentiate between what's real & what's not.💖💖💖

[@Violetikas]

Every time I watch Karolina's content I am amazed at the depth of thought and consideration, but today I am applauding while standing. This should be shown to every politic ever. Bravo.

[@OctopusOwl]

This was such a great breakdown, and Karolina highlighted many aspects of why I am frustrated with AI. A great idea, executed by exploiting millions, used to create unnecessary and harmful things.

[@CM-pf1xc]

What scares me is the amount of audio visual and film manipulation happening already, and yes people having no discernment on what’s true or trustworthy

[@madamemiu20577]

One of the reasons why i'm not spending as much time in the internet as i used to. Don't want to spend my time on fakes. I'd rather visit one of the local archives or libraries for information or communicate with my family and friends. If your life ends after switching the wifi off you've never had one in the first place. Ai just shoved this simple truth in our faces one more time.

[@matheussanthiago9685]

Fr If this trend continues, AI might accomplish what nothing has yet It'll make me touch grass

[@martinamiau6069]

YES THIS

[@vickywitton1008]

❤

[@Lonovavir]

I'm a WW 2 reenactor and can confirm the perpetuation of myths thanks to AI deep fakes and Netflix is a serious problem. People like me who LARP for kicks can't compete with the raw speed of bad media online.

[@madamemiu20577]

@@matheussanthiago9685 yeah, i actually became less depressed. People, some daily activities, playing with my cat, drawing, studying, reading and some romance related stuff became much more interesting and exciting, but the transition between addiction and freedom was hard i admit it. I was sad that i can't show my artworks online without a risk of being copied or stolen but i realised how little i care about validation from bots on instagram or youtube. I's much more rewarding to share art offline with strangers or professionals in art classes.

[@Dragon1276]

We’ve passed the Information Age and we are well into the Age of Wish Fulfilment. I can’t wait to watch my personal cut of Roman Holiday with Cary Grant instead of Gregory Peck.

[@angrytrees7519]

There very much needs to be a spoken and written disclaimer on all AI generated media. It will very quickly devolve into being on everything though, but at least we'll know. I'm imagining it being like the labels on carcinogenic food in California. It's on every restaurant.

[@nonakabyrd5759]

Darling, huge fan, 62 years old. Although I realize what your saying that changes in technology comes so quickly, there is a easy choice to ask someone knows better of a program or device. However, I was fortunate, my 2 sons, taught me to learn to manage my programs and devices. Don't forget, it was Boomers who brought the home computer to the masses. I love learning cutting edge programs and software. Also, I've met so many young people who don't know how to navigate to research. Stupid and Lazy people are of all age groups❤❤❤❤

[@sinxenon3181]

agree

[@peglamphier4745]

I'm a 65 year old professor and I'm waaaay more skeptical and information literate then my students. For a lot of them all info found on their phone has equal weight. They can't or won't tell the nonsense from the real.

[@xXNekou]

As an artist: I think AI art can look amazing sometimes, even breathtaking. But the second I realise that it was made by AI - the magic is gone, it looks soulless, emotionless, fake, because I know this was not crafted out of love or passion, it was simply generated by a robot in a matter of seconds or miliseconds, so it stops being impressive.

[@kirabad-artist6532]

And the most important, it was cobbled together from the work of other artists

[@pandapuffzee8255]

Its good reference FOR an artist. It itself is not an artist.

[@yanied9646]

There have been people that just throw in AI art to try and speed run art contests smh

[@thirdeye4654]

​@@kirabad-artist6532This is a complex topic though. It's not like any human artist really re-invents the wheel. And diffusion models don't just stitch things together at all. Everything is a remix. But most artists create things from emotions and visions. AI is missing these. For now at least.

[@ReallyGoodName3000]

Not really sure you know how AI art is made

[@blueberry_lemon]

It sucks how it's become harder to look things up online! I love learning about history, so I'm glad that there are people like you, who I can trust to care about researching and about what you put out into the world.

[@TD-ty1iq]

AI professor here... THANK YOU 🙏 for a) a historical background of the AI field b) a very SANE and on point critique on generative visual AI.

[@cybersandoval]

Antique dealers have become wise; random old photos used to be everywhere and priced to sell; now anything with historical costumes or tech or architecture are difficult to find

[@merriclearr]

that thing you said about heavy research seeping through easter eggs in movies is so true!! there's nothing i like more in a movie than noticing the tiny details that show how much they researched for it. for example the other day i watched impromptu, and all the details are so sweet, like that scene where george sand sends chopin a bouquet of flowers and he's so happy, and he really did like flowers in real life. wonderful video as always, you're the best, i love your content!! 💗💗

[@karsaanita]

Just two small additions to the AI has always existed point: 1. AI did get way more advanced in recent years. Mainly because we now have more computational power and more data than ever before. 2. The use of the term AI has been extended a few times recently because it's a hot term right now and everyone wants to get money for their projects even if they're applying "AI" technology that existed 20 years ago. Basically, we call almost everything a computer does AI.

[@AnnekeOosterink]

Yeah, especially the second part, it's a very fancy sorting program. Which is not actually AI.

[@mai_komagata]

I think of how the British Empire in Victorian times (150 years ago!) created so many myths about the past before that and the world outside of british influence and how those myths are *still taught* today and it takes so much scholarship to undo that. Imagine what the myths taught today and our overwriting of history with ai might do...

[@AugustTaylor-i7n]

This!

[@floatingdaisy3256]

Thank you for this! I’m teaching a class on the ethics of digital technology, and you raised some really good points in this video. I hadn’t considered the loops of self-citation that can happen when A.I. cites sources that were written by A.I. The point you made about people using artificial photos to discredit the veracity of historical events was really, really poignant and powerful and something that never would have occurred to me. I knew people growing up who thought that the moon-landing photos were fake.

[@friedeyeball]

I immediately want to say, thank you so much for that intro. I've been hired by a government/nonprofit collaboration to work on an educational game about AI and I've had the privilege of being able to learn about it from experts in their fields. The way AI has suddenly been rebranded is so harmful. It's like the NFT grifters came in and co-opted a legitimate field of tech. Now they're saying AI can do things where, if you understand how generative and LLM AI function, they will *literally never be able* to do what these grifters are promising. The technology underlying LLMs and generative AI are indeed revolutionary, but they're not *that* revolutionary. Grifters are selling it like the AI we have is already general AI. It's not, and being unaware of how the AI works is really leading a lot of people to make unsound financial decisions. So thank you for spreading proper awareness of the history of AI and how it's really not "new."

[@wickedthing6068]

This is fascinating. Can you elaborate on how grifters lie about AI being more revolutionary than it actually is?

[@margodphd]

​@@wickedthing6068There's good wired video with an expert that explains in depth how. Always look for your own answers if you want truth..In short ..Basically LLM can't tell truth from lie, they don't give answers to questions, they predict most likely combinations of words in context, it's auto-fill on steroids. It's same with images - it's just combining most popular images fitting the context and manipulating them to fit the prompt. It can never be correct because it's unable to know what "truth" is or isn't.

[@user-gu9yq5sj7c]

@@wickedthing6068 Shad from Knights Watch overly hyped that ai art is real art and that he was a real artist if he used that. Instead ai art is just stealing art and mixing it together through word prompts. Yt-ers like Kirkpattiecake criticized Shad and ai art about that.

[@wayIess]

​​@@user-gu9yq5sj7c Sounds like Shad doesn't realize ai art can't be copyrighted and he legally can't even have rights to his 'art'. 🙄

[@Nemamka]

The first time I heard "it's still _just_ automation, clever and complex and fast, maybe, but just _automation_ " was what really gave me perspective on AI. There is still nothing intelligent about it, it's just code reading code - words, pixels, shapes - reproducing it in different patterns. If you gave a monkey a typewriter and a million years it might just write Shakespeare's Hamlet, that sort of thing.

[@RainbowUnicorn.666]

This video is so good and sums up a lot of my thoughts on AI generated content. Whenever I dive into a mess of youtube shorts I feel like I'm surrounded by videos created by bots and commented by bots. Not long time ago I was puzzled by one (of many it appears) Instagram profiles posting quirky, old looking, victorian era photos until I started noticing that there is something way off in every single one of them and finally finding out that these are AI generated/enhanced, many times not tagged in picture's description, though there is an info about use of an AI in profile info. But what puts me off even more is amount of comments either praising it or trying to give a context to these photos as they'd be real. We definitely need a clear tagging of AI generated content. On a side note, what an AI is doing to us is what we were doing to the history, juggling narratives to fit power's narrative and hoping that people won't look to deep into it, except that we are falling into this trap willingly since it's visually more attractive than truth.

[@Shamanized]

Yeah it's that weird balance of not freaking out because every new technology made people think it was going to end the world in some way but you nicely spelled out the valid concerns that we're actually seeing right now that is definitely a problem. We've solved a lot of this stuff through education but who is going to educate those out of the loop on this tech? Policy is insanely slow on this it's pretty frustrating.

[@Sweetgotham]

You arguments about how governments would use AI is spot on. I am reminded of a great cook called "The Commissar Vanishes: The Falsification of Photographs and Art in Stalin’s Russia" which documents how the Soviet propaganda machine would 'erase' people who had fallen out of favor from photographs or to bolster positive, but fake images.

[@alextheasparagus6675]

like, i wouldn't have been able to immediately tell that the picture in 11:40 was ai if you hadn't told us, and that's really scary. when even top ai scientists, like max tegmark, are worried about the advancements of ai i think it's clear that it is a real concern and should be regulated

[@maryeckel9682]

I recently questioned that very image in a vintage fashion group I mod on Facebook. I couldn't track it down to a source, but it's off. Her dress is all wrong. She's too lean and too tall. PDA was not a thing. It's uncanny valley. But I couldn't prove it, so it stated, and it got shared. I hate it.

[@poe8910]

i’m by no means an expert on historical fashion but i think the lower part of her dress is very clearly inspired by modern “fairy-like” dresses. i used to look at the photographs from victorian era a lot and i don’t remember ever seeing that kind of fluffy and thin material as an over layer on any dress from that era, especially with that modern looking embroidery. all of it looks like a modern mesh material made out of polyester and the embroidery looks like it was made by a machine. but again, i’m not an expert so i can be completely wrong.

[@songindarkness]

It looks like it incorporated some modern wedding photos and poses. But I don’t think I would not have realised it as an AI image without being told. We are too used to believing in the evidence of our own eyes. This is very scary.

[@RiotFortuneee]

As a person whose lively hood has been and will continue to be affected by image generation I really appreciate you talking about this subject. Even just looking up pictures of animals for illustration reference has meant running into plenty of inauthentic AI trash. Its almost sad how useless google image search is right now.

[@TiffyVella1]

I am seeing this discussion cropping up more and more online across multiple platforms. Its becoming increasingly hard to rely upon google image search for reference material.

[@Tadicuslegion78]

"Is AI ruining..." Yes, yes to everything, AI should never have been allowed out of Pandora's box. It will be our doom.

[@shroomyk]

Agreed

[@Vanbedda]

Recently I wanted to start making Librivox recordings so I went to the subreddit to see if they have any suggestions and the first things I see is "How do Librivox readers feel about their voices being used for AI porn?" OMFG noooo

[@johannageisel5390]

AI in and of itself would not be a problem. AI being controlled by greedy assholes within a capitalist society is a recipe for disaster, though.

[@matheussanthiago9685]

But hey, have you considered that line goes up = good

[@serisaurusrex]

@@Vanbedda Omg I've recorded for Librivox... I didn't know this was happening!

[@andromeda_lw]

I've recently been on a historical photography exhibition in Cracow and I 100% felt what you mentioned. Knowing that these are actual people with their own hopes, dreams and stories... It's a completely different experience. What moved me especially is how much quicker people aged. Really shows you what kind of different life it was. AI could never.

[@cynthiamurphy3669]

I'm 70 years old and am pretty confident I can recognize AI stuff when I see it. I feel most people my age who spend time on the net would probably say the same. I'm not a facebook fan. I very much dislike the you tube channels that try to make all recognizable subjects of portrait paintings from the past into more attractive people, lol. I'm okay with cleaning up and colorizing old films, though. I absolutely hate cell phones and use a dumbed-down android and took off all the apps. I long for the days when phones were attached to the wall, lol.

[@kennethhymes9734]

I love your channel. And great video. One quibble. There are many historical facts and ideas which you cannot check the validity of with a click. We tend to think of the problem dealt with here: Not everything on the internet is true. But we forget that not everything true is on the internet. I will give one example: Stu Horn was a prominent collage and experimental artist in NYC and Philadelphia in the 1970s and 1980s. He wrote for Sesame Street. He had a music EP. He was socially white but discovered later in life his family hid a black grandparent from him. He had a tremendous impact on a wide range of people. And none of what I just said is in one place, most of it is not online at all. If we consider the internet the arbiter, we will eventually lose everything some doofus algorithm doesn't think will sell ads.

[@auntiegravity7713]

Yes, this is long. I know. I enjoy this channel and I’m subscribed. One of the reasons is because of the intelligent explorations of topics I’ve never thought of. Connections, things to geek out on, and such cool, fun, and original content. Thank you! I do want to point out something that bugs me (in the same way that misrepresentation of history is a pet peeve for many) The notion that “middle aged, older people, people over 50, etc are inherently “not good at using technology.” Or Gullible. As a GenXer, I'm gonna jump right into the debate and point out some flaws in this specific theory. 1. Technology isn’t foreign to us. We’ve been through this before. Many of us were building websites and were early adopters as we came of age. I didn’t go from being (somewhat) tech-savvy to becoming clueless the day I turned 50. I use the same tools and have had to keep up and adapt just like everyone else. Often, I'm the one vetting out new software and helping clients automate their workflow. I know plenty of women my age who are pretty comfortable with "tech." But AI IS humbling. Very much so. As a generation, we also understood the initial optimism and novelty when the internet was born, and not too long after, started to fully understand what it meant, the changes it would bring, and how to adapt. The theory that we’re more “gullible” as a group doesn’t hold up. 2. We are familiar with, grew up with, and conducted our studies and research with real images. We’re also familiar with image editing and quite possibly (probably?) have a better grasp of what humans really look like. We know that women’s butts don’t really look like those pictures we see on Instagram. We can certainly tell the difference between someone who is using filters and who isn’t. I'm thinking yassification will eventually die out and they we'll be the hip ones again.. who knows. We can recognize the difference between CGI and practical effects in movies. Through exposure to more experiences in "real" life over time, we just may be able to be better equipped to tell if a photo has been set up or manipulated or was taken as is. We recognize the unique sounds and nuances of a vocal performance in the days before AutoTune, when everyone didn't sound the same. I laugh at the ads that assume I can't tell the difference between a voice actor or an AI generated voice. It's very obvious. (Although this could change) We studied HISTORY. :) And of course we're exposed to new developments as well, just like everyone else. Having a base understanding of reality while also having grown up with technology can actually be an advantage. 3. Lazy thinking/making assumptions about age: A lot of younger people seem to be stuck on this notion that someone who is 50ish had the same experiences as your average 80 year old. It sucks to be lumped into this category. Or that being 60 10 years from now will be the same as it was for people 50 years ago. 3. I would agree that on Facebook in particular, there’s a very interesting display of the interactions between.. well, not the brightest people on the planet. Although there are communities (I’m an admin of one) that cater to the more educated within a specific niche, overall I’m constantly amazed not only by the lack of intelligent and nuanced discourse, but how catering to the lowest common denominator is almost encouraged. (So-called “community standards) Some of the dumbest things come from big groups.. including the GenX groups. However, I do see a lot of stupid crap on Instagram as well. It just has its own brand of stupid. 4. Yes, conspiracy theorists can be duped by fabricated images. However, sometimes people that are assumed to be conspiracy theorists turn out to be correct because the DON’T believe everything they see or hear. The emerging technology can also be used to produce propaganda on a wider scale, and resisting this could be difficult or even dangerous. (I think this was pointed out) Could you say that THIS is a trait of GenX? I think so. As a group, we can be pretty cynical. Gullible? Nope. Being older also generally means more life experience, which fuels a healthy skepticism. Example: . I happen to live in Bosnia at the moment. Although I did not experience the war firsthand, you bet there is overt and subtle evidence of it soaked into every aspect of life here. I can see how fabrication of fake events could be used as a weapon, so you bet I'm going to be even more careful about what I see and hear about just about any topic, let alone this one. Edit: I'm gathering that Karolina also has some knowledge and experience with parts of Europe that have been through a lot .. a lot of people in my native country (US) have not. However, the older they are, the more likely they have more experiences outside the US and outside of what they are used to on a daily basis. We need better stewards of history. My point is that yeah, with AI on the scene.. better that we have people around that have wisdom, experience, and accurate memories.. or are at least remaining skeptical about what they see. 5. The explosion in AI technology and its implications is new to ALL of us. As was emphasized in the video, the implications are going to be intense and deep. There will be ethical, legal, moral, and creative questions to explore that even experts (lawyers, SEO experts, the list goes on) are barely equipped to deal with. I think it's incorrect to assume that a young person, just because they are young..are automatically gifted with knowing more than a copyright lawyer or marketing expert or professional musician or technology experts about these implications.. and the ones I trust also acknowledge that they can only do so much right now. (humility) OR that they possess more finely-tuned AI/BS detectors. In fact, one could argue the opposite could be true in many cases. (Of course not all) To assume that younger generations who were raised with mobile devices have an advantage over GenX doesn’t compute. Knowing how to use tech as a consumer isn’t the same thing as developing it or applying it to a particular discipline or creating/using it for a living. I’ve seen a LOT of younger people that only passively use technology as a consumer, and still seem clueless about it. 6. I’m not a fan of ageism in tech. This is something I speak up about, much like Karolina’s peeves about the interpretation of history, so I get it. It’s good to speak up about things like this. It’s not about us being “sheltered” from truth, but debunking false myths about what it means not only to be a certain age, but part of a different time and generation than our parents and grandparents. I wanted to mention these things as part of a nuanced discussion, of course. One of the reasons I enjoy watching my favorite KZbinrs is because it encourages people to think (while also being entertained) There is a lot to unpack with AI, and it can be tempting to resort to lazy thinking, much like what people do with historical interpretation and costume design. Soo.. this video made me think, and about how much we also agree on! I love this channel and it’s perfect blend of information and entertainment. You got the “Trad Goth” look down (although there are variations on this theme.) but I’m also starting to understand more and more that there is a connection between what we wore and the experiences we’ve had, and how we are products of the times and environments we lived in. And that it's important to retain some sort of accurate documentation of history, including what people wore. Thanks for all you do. It’s not easy being a KZbinr!

[@k80_]

I think your assessment at the end of the capital nature of all of this is very apt. Even the existence of these programs on a higher level is such a capitalist invention. Produce infinite product without any overhead. It’s kinda chilling in that light

[@1945iwo]

As a military Historian, I developed a new pet-peeve with AI not getting the uniforms right at all correct.

[@bluedragontoybash2463]

gen Z ? millenial ? because gen X have trouble shifting through internet fact

[@EFO841]

one of my favorite videos that I still come back to is that "100 years of fashion in 100 seconds" from like 2012! they did a montage of different fashions but with real people dancing to different music styles from the time as well. much better than that instagram reel

[@uwatenage9]

Karolina, i have watched your videos for many years and really enjoyed them (being a historical reenactor myself), but this video stands out for it's exceedingly intelligent perspective and salient discussion topics raised. great stuff, seriously. super fanboy over here. carry on.

[@manouckgermain8945]

I totally agree with your point of view. I used to be annoyed when there were an imperfection in an image or even live concerts recordings. Now I seek the presence of those imperfections because, in a strange way, they reveal the harmony or beauty that surrounds them. Weird ? Thank you for your input ( no emoji).

[@AdeleiTeillana]

I agree with most of what you said. The only thing I take exception to is your comment about Gen X and older generations being more easily fooled by stuff on the internet. That's a misconception, a pretty widespread one, but in the spirit of discussing things which are fake versus things which are real, I feel like it needs to be mentioned. There have been multiple studies lately which show that Gen Z and Gen Alpha are actually just as, if not more (in some studies) likely to be fooled by fake news or content than older generations. Similarly, Gen Z is actually more likely to be scammed than the elderly. The reason you don't hear about it much is that Gen Z is losing smaller amounts of money, because they don't have as much money yet, still being relatively young. But they're actually more susceptible to falling for scams. (They also can't blame dementia!) Anyway, I'm not a member of any of the generations I mentioned, so I've got no skin in the game, so to speak. I just think it's important for younger people to realize that they're just as vulnerable to falling for fake information or scams as anyone else is.

[@Ceruleansquid-lo3iv]

I think that Gen Z usually knows about scams like the Nigerian prince, and that's contributed to the idea. They don't fall for the famous and obvious scams, but that doesn't mean they don't fall for any scams.

[@MarieAnne.]

@@Ceruleansquid-lo3iv Almost everyone knows about the Nigerian prince scam. I'm 63, and that scam's been around since before the internet.

[@jefferystube]

I have begun to think of what ai is doing to the Internet as "Content Collapse Disorder." Like what you described with ai based on Wikipedia, a sort of degradation will happen when ai begins to source data that was also ai generated. We also already have ai content being liked, shared and commented on by bots. A portion of content being created artificially, a portion of the content being "consumed" artificially, edging out creators, and lowering the value of genuine human content.

[@songindarkness]

I always come back to a phrase I came up with a while ago: content is nothing without context. We are seeing now what happens when we have an unimaginable amount of content and not enough context. We have content divorced from reality in endless loops of impossibilities with no imperfections. I very much agree with your statement “Content collapse disorder”, but maybe it’s more of a “context collapse disorder”.

[@TheLaughingPanda]

This is a really great video, I can tell you've really put a lot of thought into this! For me, my main problem is that AI images have the same problem as the AI text generation, which is that the AI is just a "generator". If you ask AI to write your essay, it will invent sources. If you ask AI to make something historical, it will invent history. It can make things that are nice to look at, and if training data was gathered with the consent of the artists I don't think it would be ethically any different from photoshop, but both the creators and the viewers need to be aware that it only creates "fantasy" images, and not try to pass off AI generated images as in any way "true" or "real".

[@tirsden]

I'm gonna pause the video and say that you've been a very big influence in my perception of fashion and lifestyle history, and I want to thank you for that. There are other informative influences elsewhere, but as you were listing off the cliché "first image associations" for various time periods, I didn't actually see the things you described, and a huge part of that is literally thanks to you. Ten years ago, I definitely would have seen those stereotypical images in my head, though less so for eras I studied heavily as a teen and young adult (like WWI and WWII because uhh, someone used to be a huge Indiana Jones simp).

[@MissMisnomer_]

We have officially entered the Gaslight era baby

[@Unknown2Yoo]

When it comes to seeing the past, that's why I'm so happy I work in an archive. I'm an artist, and amateur photographer, so being able to work with photos and images almost all the time is so rewarding and educating. Being able to know how things are "supposed" to look, is invaluable. I encourage anyone and everyone to investigate museums in their area, or online, to have access to their digitized or physical collections whenever you want to do research. As technology is becoming more easily accessible, it's been easier to get more and more images online for non-profit entities, so there's a lot of material out there to sift through!

[@RowieSundog]

I struggle with some mental disabilities and I absolutely hate how now I have to seconds guess every single image I see

[@JohnDRuddyMannyMan]

I’ve started running into this problem more and more when I’m researching my historical illustrations. As if secondary resources weren’t tedious enough for their accuracy, AI doesn’t even know that it doesn’t care about historical accuracy!

[@audriscarborough9309]

Katerina, I am very grateful you have opened the dialog about this!! I am afraid I will go on a tangent... So, I just want you to know that I am agreeing whole heartedly with what you are saying. I have the same fears and concerns. I absolutely think we ( my Country of United States of America) and other Countries, need to stand up right now and lay down some rules, regulations and an absolute way to tell, real from fake. As well as lay out some HUGE and ENFORCEABLE as well as ENFORCED ( because we all no how we ignore and let slide some incredibly horrific crimes - if it suits the people making money!! Now where do we go to get on board with this?? Obviously, we need to take the time to search that information out for ourselves and then take a bit of our time and do something about it!! Thank you for speaking truth to what is happening. Be well and happy, Audri

[@aminadoce]

Yeah, this one looks like AI lol

[@isamukim1693]

«But I dont judge...» with the absolutest shady face 😂

[@marinalodi6313]

AI should have an obligatory watermark indicating it is AI, or any type of disclaimer

[@ThulrVOdinsson]

A.I. "art" usually has this unnerving, formulaic, lifeless feel to it. This is a wonderful video! We need more people speaking to this. I'm no doom-speaker, but there are definitely concerns with A.I. that we really should stay on top of, and I do fear we could drown in A.I. misinformation within the near future if we're not careful.

[@fakehistoryhunter]

Thanks for making this video, I've been debunking AI nonsense on twitter & facebook for ages and wrote about it in my book, all we can do is make people more aware and suspicious of what is thrown at them. It's a fight we historians and history lovers have to deal with more and more.

[@NathanDudani]

debunking lol

[@fakehistoryhunter]

@@NathanDudani Yes, literally.

[@ajbryan59]

“Every record has been destroyed or falsified, every book rewritten, every picture has been repainted, every statue and street building has been renamed, every date has been altered. And the process is continuing day by day and minute by minute. History has stopped. Nothing exists except an endless present in which the Party is always right.” ― George Orwell, 1984

[@lfgifu296]

there’s one thing I’ll thank AI for and that’s the US presidents play Minecraft videos💀

[@f0restangel]

THESE WERE SO FUNNY LMFAO

[@lfgifu296]

@@f0restangel they were the peak of XXI century comedy istg

[@ThinWhiteAxe]

YES

[@matheussanthiago9685]

​@@f0restangel they were cool for about a week until everyone and their mother started uploading the same uninspired president AI gameplays

[@aerialpunk]

Also, those videos where they put Arnold Schwarzenegger's face on classic movie characters and read the lines in his voice (and same with musicians).

[@electricbrighto625]

Your messege at the end was really powerful. Thank you 👏

[@braddo7270]

You lost me at "stolen wealth". You're a HISTORIAN, talking about losing a grip on history... then you say THAT? Stolen? Really? Care to retract that statement? 👀

[@WitchKing-Of-Angmar]

You'll find that an right leaning historician, a nazi historician, and a leftist historician tell history in three different ways. Bias is the key to loosing touch on the reality of the past. She stated that there was a movement going on the entire time and that the fancy dressed lady leaving the shopping store of choice was getting on a segregated bus. But women had the right to vote in 15 states before 1920 in America, and she looks definitely like she's in New York, and there wouldn't even be a town style bus in the city, let alone a segregated bus up north. Maybe a virginia bus which wouldn't make any sense, they'd have to switch their rules.

[@I_am_Lauren]

God I love you karolina! Never stop talking about these subjects! We are the same age and I got a history degree so your content is amazing! As someone who does research for a living it is frusterating and baffling that people don't know how to do a basic google search and know to use Primary Resources.

[@patricequinn-ee1iy]

Ooof, this reminds me of two decades ago, myself and film director husband at a weekend conference of filmmakers discussing high def. You are the answer we should’ve known- ourselves; storytellers. Historians, teachers, students; friendly, passionately so; members of life. You teach, remind, inspire. The only answer is to give it. Thanks, girlie ✨

[@Artevibe]

This is my favorite video you've ever made. So eye-opening.

[@agnieszkamalicka7232]

Nie myślałam o IA jako narzędziu do zakłamywania historii, ciekawy film, ważne kwestie poruszone. Na pewno będę się bacznie przyglądać obrazom z historii od teraz 😅

[@tamarakonczal6350]

Man. Everything is history. Oh Karolinka. Glad you are alive.

[@elizabethtaylor3600]

Another timely and fascinating video, thank you for sharing!! Your content keeps me coming back to KZbin

[@oliviakot7972]

I would also add that the recent history, like 80s, 90s, early 2000s are getting completely wiped and replaced with enhanced images. The amount of times I've seen takes from movies or pictures of celebrities from the time, enhanced with AI filters to fit the modern beauty standard... And they are quite believable sometimes!

[@thomasdamours7325]

THANK YOU. I've been simmering about this so much over the past few months. looking up historical styles and stuff on image databases has always been an exercise in discernment, but now more than ever it's so difficult to just. Filter out the trash and clutter and visual misinformation because AI can mimic authentic visuals in a way which is very hard for humans to do, and it complexifies so much the process of sorting out the real from the fake. And it's frustrating, and worrisome, and as someone who cares about studying and sharing history, it feels like the forces working against us--and were already powerful--just got 10 times stronger. I just really hope we can figure out ways to oppose this process effectively.

[@AtlasAdratica]

I actually did a paper just recently on how inaccurate historical media claiming to be accurate is really detrimental to modern education about the past. Makes sense AI would play a role as well. Awesome video it was an viewpoint i hadn't even considered!

[@riveranalyse]

A huge percentage of those disturbing likes and comments will themselves be bots. Which is something of a relief in one sense and another nightmare at the same time.

[@carolineroper5509]

I just came on here to say that I'm a professional data scientist with published research in LLMs and I think everything you said makes sense, and this was a pretty balanced and thoughtful take. Appreciate your perspective!

[@savingplayer1613]

She ate and left no crumbs! Go off, girl!

[@grantking2880]

1. Thank you for talking about this Karolina. I've enjoyed watching your videos for a long time, and I especially appreciate your perspective on this topic. This is a very important topic that is close to my heart. 2. I get what you are saying. I hope the people who are working on these models will watch this video and be inspired to engineer new guiderails to improve the historical accuracy of output that is meant to be historically accurate. I sense that you already get this, so just the questions in sections 5 and 6 and 7 are for you. 3. For people who aren't up to date with the use cases of AI, I think it is important to emphasize that as of today, consumer AI itself isn't able to autonomously post or promote content. Custom solutions that can do that are highly experimental and require complex software architectures that orchestrate much more that a single AI model. If a human operator is employing one of these experimental orchestrations, they have a moral (and usually a complex legal) obligation to explain everything to the consumers of that system's content. 4. I'm saying all this because it I believe is important to make a clear distinction between a specific tool and the result that its wielder creates. At this point consumer generative AI is a productivity enhancement tool. 5. With greater accessibility comes greater incidence of misuse. Considering the overall benefit for humanity, do you think the benefit of greater AI accessibility justifies the likely inevitability of a higher incidence of AI misuse? One example of misuse would be the human who used generative AI to create the components for the historically inaccurate and misleading content. 6. If another creator used the same tools and more thoughtfulness to produce something that was historically perfect, would you make a video with a different response? 7. Would you be willing to put your thoughtfulness and expertise behind the productivity enhancing powers of generative AI to compose a historically accurate history of fashion? Would you consider using AI to help you come up with the most effective image prompts? Thank you for your videos. I respect your thoughtful work and dedication to accuracy. If you ever touch on the topic again, I'll look forward to hearing more.

[@mikeharrison1868]

This is genuinely terrifying. The people who manage to control the past will control the present, and thence the future. Every AI image should be stamped, internally at least, with its creation details.

[@BethAge95]

As an already anxious historian this makes me even more anxious... Alternative history and conspiracy theories about history are already prevalent, but if these people are able to create images to prove their "truth" it's gonna be so much worse. Another sub chapter for my phd on historiography to research and write...

[@sebastianho447]

This opened my eyes after you explained to me from your specialist niche. I'm also afraid that we are having more tools to create a history of the past that suits our needs...

[@Elyanon]

Long time subscriber and first time commenter. I really love this video. You always present your factual research in a fun and engaging way. You said everything i have been feeling recently. Thank you for your work and giving us such quality content.