RE: The Steem News @ 14 March 2023 : Community Curators for April - Applications Open...
I realize that the use of AI technology in the context of Chatbot / ChatGPT has had a negative impact on the platform and as a verifier in Steem For Betterlife Community has also emphasized and emphasized that it will not tolerate attempts to abuse it.
ChatGPT dynamics take a long time to detect, I see from several tools used, not always finding a satisfactory level of accuracy.
Case in point, I made an experiment on detecting the use of Chatbot which I wrote manually in my own language and then translated into English, the article was detected as writing by AI, maybe the translation from the original language was considered as not human writing, was it because the grammar was inappropriate or grammar.
Then I tried using the Chatbot service and the results I pasted into the GPT Detector, surprisingly even the text was considered to be written by humans.
In my view, the use of tools to detect AI writing must be done by utilizing many tools, this is to compare the results of AI writing detection.
Here are some tools that I use to detect, either web-based or add-ons:
- https://gptzero.me/
- https://corrector.app/ai-content-detector/
- https://openai-openai-detector.hf.space/
- https://writer.com/ai-content-detector/
- https://sapling.ai/utilities/ai-content-detector
- https://copyleaks.com/ai-content-detector (add-ons for google chrome)
- https://contentatscale.ai/ai-content-detector/
- https://huggingface.co/spaces/PirateXX/AI-Content-Detector
Of the 8 tools above, I often use points 3, 6 and 7 as the main.) Hopefully, this can be a reference for users in detecting AI content
That is a good list of detection tools.
ChatGPT will present many problems for Steem.
Accurately detecting AI generated content is very difficult.
Excellent! Thank you!!!