It struck me how seamlessly AI can now mimic human creativity, to the point where it’s nearly impossible to distinguish between human-made and AI-generated content.
And let any non trained people to subjectively judge if the content is AI generated? With the latest versions of video generation models we are already in the situation where only software can tell if some videos are AI generated or not. Lots of channels will be wrongly tagged.
Fair enough. Though I don’t know about your point regarding the software tools. The ones I tried had pretty bad accuracy (and were quite resource-intensive to run)
So, I felt like the human-tagging way was still the most sensible way forward. At least for now, where there are still subtle cues that give AI-content away once in a while.
And let any non trained people to subjectively judge if the content is AI generated? With the latest versions of video generation models we are already in the situation where only software can tell if some videos are AI generated or not. Lots of channels will be wrongly tagged.
Fair enough. Though I don’t know about your point regarding the software tools. The ones I tried had pretty bad accuracy (and were quite resource-intensive to run) So, I felt like the human-tagging way was still the most sensible way forward. At least for now, where there are still subtle cues that give AI-content away once in a while.
Adding a link to such a tool with “If you are unsure you can use xyz. Note that these are often inaccurate” would be enough to address this.