I don’t really look at images to see if they might be AI. In many cases that are pointed out they seem pretty good to me. But this incredibly low effort caught my attention. The whole image is pasted up like a beginner using a clip art program. Especially the right hand drive 1971 BMW Bavaria cruising more or less across the desert. Yeah, there’s a relevant car. Although I do wish I still had mine!
I sort of feel like there is, or soon will be an entire category of ironic AI generated images that are memorable specifically because of their surreal badness.
I mean, that image is awful, but it’s awful in a specific, weirdly on-topic way for the story it’s associated with.
Journalism degrades as copy editors hunt for cheaper stock images
I never really thought of stock photos as essential to journalism so much as “fills space and gives person something to attract the eye to the story”.
There’s a sushi place near me that’s quite good but has some hilarious AI posters in the window. I’m less concerned that they were made with AI and more that they were made with some 2nd generation AI that’s advertising “RAMZN” and “RAMEIN: Raissamiin R Aanin Roin”
It’s like those posters for the Wonka fiasco that promised “Woolnd of Purze Magginaggtion” and how people could have thought "This looks legit.
Honestly, that picture looks exactly like a crappy Photoshop effort. No need to pin it on AI.
Also, FWIW, that’s not an ad; that’s an illustration which was apparently made for the article. It credits “Getty Images,” so I’m guessing that someone put the image together using a bunch of images from Getty.
That isn‘t generated by AI; that is a cheap-looking collage in the style of a ‘Fifties travelogue.
AI-generated graphic and photographic art has a distinctively odd and offputiting style; proportions, details, shading and lighting,and stylistic consistency (for hand-drawn images) are absent, and they all have a vaguely oversaturated look reminiscent of a Thomas Kinkade painting.
That “Ramein” poster looks like a Cthuloid horror emerging from the bowl to attack the customer and I’m not even sure what the green vegetable or the vaguely citrus-looking slice in the bowl is supposed to be. Neither look appealing and I’d be inclined to pass on a Ramen restaurant with those in the window because apparently the owner doesn’t care to advertise in a way that represents an appealing product.
Stranger
And the art director, Justine Goode, is credited as well. No one is pretending that’s supposed to be real. Newspapers and magazines have used such illustrations for decades, even before the rise of AI.
100% agree this is a retro collage, assembled from stock and other photos. Zero AI here.
That first picture looks exactly like an ad for a religion to compete with the Flying Spaghetti Monster. Ramenfarians.
There is this woman who owns a barbershop in town. She uses Facebook as her shop’s website and she used to make these crazy, rambling posts about how people just don’t care anymore, her lung infection is back, her car broke down: anyway the shop is open 11-7, come on down.
Someone must have told her to tone down the crazy, now she posts the shop hours under these AI generated photos and people now complain on her feed that her cinderblock building looks nothing like the beautiful photos she posted.
There’s an online shop I was browsing for various imported foods - then I noticed they were using AI generated images for the products on sale - such as this one for a babybel cheese product:
I’m fairly sure Babybel doesn’t make cheese juice, or weird cheese bears or wall-mounted sunshine cheese sculptures.
Or this one:
It’s Monster Munch, not Munster Munch. Also Disney/Pixar’s lawyers are waiting on line 1.
Or this one, with some duck portions that can’t decide whether to be legs or whole ducks, and a bottle of duck confit in the background:
And they also don’t make Swiss cheese with holes (I know, they used to be my client).
It’s mainly a sushi joint and quite good. I saw the posters more as the owners (or kids of the owners) saying “You could make some neat posters for cheap!” than an indictment on the place. Amused me more than anything else. If they had left the text off, they would have been “better” (more boring)
In the grim future of Hello Kitty, there is only Babybel.
Stranger
Yeah, there’s only like one or maybe two items in that picture that really resemble their products.
There’s a pub in my village that uses AI generated images on all of their facebook posts promoting various events like the pub quiz or a special sunday carvery or something; I can’t share any of it for privacy reasons, but they were just dreadful - clearly made with one of the models where they tell you ‘don’t ask it to generate text, because it can’t’ - there was a poster where the text says CARRVERRY Janiuary pub - and although the picture part of the image looks like an arrangement of food, when you look at individual items, none of it makes any sense - weird cubes of meat; potatoes that are shaped like mushrooms; greens that are actually fern fronds, with shrimps peeping out of the foliage a bowl of what might be peas, if they weren’t purple - I just can’t understand how anyone would have looked at it and thought “Yep! That’s what I wanted”
During the awards ceremony at my school a couple of weeks ago, there was a slideshow of various images. Most of them were pictures of the students who were receiving various awards, fairly straightforward. But near the start, there was a picture of some folks in graduation attire. Most of the students noticed that the accessories weren’t the right color for our graduation attire. What I noticed, though, was the the tassels hanging from the hats didn’t have continuity. You’d see the tassel on the top of the cap going off to the right, and then the bottom of the tassel was on the left, just dangling in midair unconnected to anything.
What gives? It’s not like they didn’t have any images lying around of our actual graduations.
That’s a deliberate collage aesthetic. Nothing AI about it.
I agree. That’s not AI. Or, if it is, it’s AI that was deliberately prompted to create a montage-type graphic. It’s not a brilliant graphic but I actually like it more than the average AI foofaraw. It looks more “real” to me (as in real news graphic).
Splitter!