Animals that do not exist are becoming real, and the behavior of existing animals is increasingly being fabricated. Anyone who spends time doomscrolling on Instagram, TikTok and other platforms may have noticed it: the animal world is now being flooded with AI videos as well. These are not just funny, mostly harmless cat clips, but videos that appear to be trying to convey knowledge, at least on the surface, meaning the serious and supposedly productive side of social media.

However, the growing number of supposed nature documentaries with dramatic effects makes this more difficult and even contributes to the spread of misinformation. Anyone who is not active in the relevant field is unlikely to notice. There are so many different living creatures on Earth that hardly anyone can know every animal and plant while also knowing how they behave.

It probably will not be long before social media users encounter new animal species that do not exist and watch behavior videos that are completely unrealistic. We have not seen fake animals yet. The latter, however, is already happening on a larger scale. One fairly viral video, for example, showed a baby whale being fed milk by its mother. The milk is shot into the calf’s mouth in a spectacular stream.

Counter-videos explain AI

That has at least led to a countermovement. Chris “GatorChris” Gillette, for example, drew attention to this whale video in an Instagram clip and corrected it. However, his chances are limited. Even with his follower base, he can reach only a fraction of the likes and therefore the reach achieved by those spreading the video. That applies both to the original creators and to Instagram’s comparatively new Repost feature, which lets the Instagram user base spread it further.

A real video showing a whale nursing, by contrast, has no chance. GatorChris referred, for example, to a video of a nursing whale by Cassie Jensen. With 2,500 likes, even among whale enthusiasts hardly anyone is likely to have seen that video.

Spectacular videos are rewarded

Moreover, Instagram, for example, does serve AI videos that are formally labeled as AI. But that is only noticeable when looking at the details. Social media users can do little about it. They sometimes do not even learn what is happening on the platform, because the platform itself has become the source of information. Once the algorithm veers toward popular content, which Instagram in particular repeatedly tries to do, it is difficult and laborious to correct. Even then, false information still reaches users.

A remedy would only be possible if social media, for example, were addressed in schools with a high level of expertise. But in Germany, even computer science is not covered adequately. Qualified specialists are in short supply. Until social media could be taught properly in all schools, decades would likely pass, as with computer science. On top of that, older groups of social media users have few opportunities for continuing education. Adult education centers tend to focus on how to place content on social media, not how to deal with it. If awareness of the problem is also lacking, people are hardly likely to think of looking for such courses.

With some luck, however, doomscrolling may also lead to scientists who convey information with fewer effects and simple means. Those accounts are best kept in view, because that changes the algorithm significantly, even if it does not prevent everything. In our experience, however, the probability of encountering such videos is extremely low, especially when likes are given only to effects-heavy clips.

For the platforms, these AI videos are also a blessing. From a purely economic perspective, there is a strong case for supporting such videos. They are often watched, liked and commented on, allowing a lot of advertising to be served. A change is therefore hardly to be expected.

It is therefore only a matter of time before animals are discussed at school, among friends or at work that should not be discussed.