A new tool aims to help AI chatbots generate more human-like text – with the help of Wikipedia’s guide to detecting AI, As reported Ars Technica. Developer Siqi Chen says he created the tool, called Humanizer, which explains the Anthropologie cloud, which Wikipedia’s volunteer editors put together as part of an initiative to combat “poorly written AI-generated content.”
Wikipedia’s guide It contains a list of signs that the text may be AI-generated, including vague attributions, promotional language such as calling something “breathtaking” and associative phrases, such as “I hope this helps!” Humanizer, which is a custom skill for Cloud Code, helps the AI assistant avoid detection by “removing signs of AI-generated writing from text, making it more natural and human.” According to its GitHub page.
The GitHub page provides some examples of how Humanizer can help the cloud detect some of these, including changing a sentence that describes a location as “settled within a breathtaking area” to “a city in the Gondar region,” as well as adjusting a vague attribute, such as “Experts believe it plays an important role” according to a 2019 survey …” Chen says the tool will “automate Will push updates “when Wikipedia’s AI-detecting guide has been updated.”
It’s only a matter of time before AI companies themselves start adjusting their chatbots according to some of these, as OpenAI has done Overuse of ChatGPT has already been addressed of the EM dash, which has become an indicator of AI content.
