Unfortunately, such detection methods are often completely unreliable. It might be able to detect obvious cases, but it’s no good at detecting the difference between smarter bots and people.
There’s a guy called upper echelon on YouTube that made a video showing how much social media is actually bots and its terrifying! Something like more than 50% aren’t human posts.
There’s no doubt we’ve seen an increase in convincing bot posts across social media. It would be interesting to apply some LLM detection methods to Reddit to see growth of bot posts over time: https://medium.com/@rxtang/the-science-of-detecting-llm-generated-texts-e816a14c18d
Unfortunately, such detection methods are often completely unreliable. It might be able to detect obvious cases, but it’s no good at detecting the difference between smarter bots and people.
There’s a guy called upper echelon on YouTube that made a video showing how much social media is actually bots and its terrifying! Something like more than 50% aren’t human posts.
Removed by mod