Tail Wagging Cats, Veterinary Implications of AI Generated Videos
Behind the Paper
I was absolutely fascinated when the Sora2/Nanobanana/Veo 3.1 iteration of generative AI models came out. Suddenly, we can have a very plausible looking video of an animal doing crazy things. I was fooled myself by a few of them!
Sora 2’s press release had so many animal-related videos in it, these were meant to be the most impressive, most compelling uses of the tool for advertising its existence, and yet these were plenty of examples of questionable behaviour or potentially harmful interactions.
In this paper, we deliberately chose to use only the videos contained in the press release, as this saved us from having to generate our own videos, and didn’t mean we were using other peoples’ content. As an exploratory first step, this seemed like a good ethical compromise when dealing with AI generated content.
Video
You can watch a short video about this paper here on LinkedIn or Here on TikTok
