I quite like Kurzgesagt as a youtube channel as it’s space themed videos are always interesting to watch, but recently they are claiming that AI slop risks ending their channel, mostly due to it being easy to create and looks good enough at a surface level to be engaged by people.
Video here:
They mentioned in this about brown dwarfs(A very small star that doesn’t undergo nuclear fusion, basically jupiter but larger) with their own personal hatred, and that the AI made up facts it sourced from an unknown data stream. Similar prompts from other people lead to an AI generated video which they[Kurzgesagt] claim will now be fed into newer models via bot scraping which is already happening with current LLM’s. This may lead to a misinformation loop which causes more slop to appear.
Likewise they also predict that the majority of internet users will see majority AI generated content soon, which would affect their own view stream due to the time required to make their videos compared to AI generation.
So, what can we learn from this, is Kurzgesagt right on this issue or is it fear mongering?
Personally I am more on the side of Kurzgesagt and they also have their own source list if you want to double check their claims, and I would rather have human made content rather than AI generated content. Also note that Kurzgesagt are not anti AI, they admit to using it as a tool among other more traditional tools, including using AI as a google supliment but always double checking sources for accuracy.
But what do you think?