The tightened situation on the eastern border has also been seen on social media. The already heated discussion is further confused by images created by artificial intelligence, some of which seem to go over well with their audience. The professor hopes people have patience and artificial intelligence literacy.
During the beginning of the week, the messaging service X has circulated photos in which light-skinned women who appear to be aid workers feed dark-skinned men cake and other pastries. The images are not real photos, they are created by artificial intelligence.
Only last week we reported on how the artificial intelligence image from the “eastern border” that was also spread in X had gone to a part of the social media community.
One of those who shared recent feeding photos is Vihti’s municipal councilor, Basic Finns Sampa Granlund.
– I understand if hoivavetti wants to get a dog, Granlund stated in his publication, which has since been added by X users with the additional note that the images are made with artificial intelligence.
Granlund’s post alone has been shared more than 70 times at the time of writing and has been liked hundreds of times. More than 200,000 people have already seen the publication.
The story continues under embedding.
If the embed is not visible, check it out here.
“Where does artificial intelligence get its model from?”
There is no certainty about the origin of the artificial intelligence images shared by Granlund, but on Monday similar images were widely distributed in the messaging service X, for example behind the nickname Keijo Kaarisade. This is a well-known parody account.
Reached by Iltalehti, Granlund says that he grabbed the photos from Kaarisatee’s account. He says he knew the pictures were made with artificial intelligence, but denies that the purpose was to spread false information on purpose.
– I’m not making any propaganda, after all, social media is now full of false images, says Granlund.
Does Granlund think that spreading such images will play into Russia’s pocket, for example, when confrontation on social media increases?
– I don’t think it’s playing in anyone’s pockets, social media is a playful platform.
Granlund later admitted in another post that the photos he shared were not authentic.
However, he asked where the artificial intelligence got its model and shared a years-old photo taken in Finland, where a woman wearing an attention vest feeds a cake to a bearded man also wearing an attention vest.
– Where does artificial intelligence get its model? I’ll leave a tip here, Granlund asked.
The story continues under embedding.
If the embed is not visible, check it out here.
As pointed out in the comments of the publication, artificial intelligence does not need ready-made concrete examples to create the requested fake images. The real picture must have served as inspiration for the user of the artificial intelligence application when creating the picture request.
Granlund: People could be sharp
We asked Granlund why he knowingly shares AI-generated images without telling them they are.
– It fit that old picture so funny. The picture was so similar, Granlund says, referring to the above-mentioned photograph, but admits that he does not know what the relationship between the people in the picture was.
For Granlund, the questions of responsibility for artificial intelligence images and their sharing are difficult. However, he does not feel that he is part of the problem by sharing the material. He puts the responsibility on the shoulders of social media.
– Yes, these commenters will definitely see my new post as well.
– It was already known [että kuvat ovat tekoälyllä tehtyjä], but not everyone reads those comments. People could be sharp.
Many of the commenters seem to swallow bait with hooks for days. Based on their reactions, others have considered the photos shared by Granlund to be genuine.
– Damn, what idiots, the basic Finns of Turku Paula Hiltunen commented on Granlund’s publication.
– Do they imagine that those guys can’t hold a spoon in their own hand? Hiltunen asks.
It was also pointed out to Hiltuse that the images in question were created by artificial intelligence.
– I realized when I read the comments. I didn’t look that closely at first, Hiltunen answered.
According to Granlund, the sharing of artificial intelligence images was basically a dispute between the right and the left.
“We are happy in Russia”
Professor of Computer Science Hannu Toivonen The University of Helsinki does not speculate on the motives for sharing such images.
– The pictures can very well be made by a single person himself, but you can also ask what is the association and information that has inspired him to make and share these pictures. In any case, Russia is happy when the refugee issue causes an uproar in Finland and divides the nation, Toivonen tells Iltalehte.
Toivonen says he can only guess the reasons why “propaganda” cannot always be questioned.
– One factor may be that people more easily accept information that corresponds to their preconceived notions.
Don’t believe everything you see online
In general, Toivonen is of the opinion that dividing images into genuine and generated images is unnecessarily black-and-white thinking. Different kinds of influence have always been done.
– Pictures have always been edited, and now they can also be edited using artificial intelligence. For example, new elements can be embedded or generated in the right image, or existing ones can be deleted, Toivonen points out.
– Using new artificial intelligence tools, generating and editing images is even easier and the results are more convincing than before. Soon we will also see more video clips.
– The old wisdom that you shouldn’t believe everything you read on the internet should be extended to touch what you see as well. You shouldn’t automatically believe everything you see on the internet.
In addition to media literacy, Toivonen also calls for artificial intelligence literacy. He thinks that recognizing artificial intelligence images will only become more difficult as technology develops.
The story continues below the picture.
This picture was also made with artificial intelligence. Be careful. Vertex Space / Adobe Stock
AI image or not?
According to Toivonen, a closer look at the image can reveal that the image was generated by artificial intelligence.
– The most obvious signs are obvious errors in image generation, such as six fingers on one hand, the merging of the legs of people sitting next to each other, nonsensical text, various discontinuities and other strange features, he lists.
– Some image generation programs produce unrealistically smoothed images. For example, the face is smooth and the smile is crowded, but they can be like that even in real pictures.
However, the image can be generated by artificial intelligence, even if it does not have any of the mentioned features.
– If you want to cheat with pictures, the scammer will of course remove such pictures, Toivonen points out.
The story continues below the picture.
The child in the picture doesn’t really exist, because even this “photo” is generated by artificial intelligence. Fotostockerspb / Adobe Stock
There is not necessarily anything in the picture that could be identified with certainty as being made by artificial intelligence. In addition to the technical characteristics of the image, Toivonen also urges to pay attention to the wider context.
– The way a picture is used can reveal a lot. Do you want to use the image to attract visitors to the website? Draw attention? Entertain? Provoke? Make fun of someone? Manipulate mental images? If the image arouses a strong reaction, you can think about the motives of the person spreading the image for generating or editing the image.
– When talking about propaganda or manipulation of mental images, scams often seem to try to smear. That’s one good sign.