Why detecting AI-generated text is so difficult (and what to do about it)

This device is OpenAI’s response to the warmth it’s gotten from educators, journalists, and others for launching ChatGPT with none methods to detect textual content it has generated. Nevertheless, it’s nonetheless very a lot a piece in progress, and it’s woefully unreliable. OpenAI says its AI textual content detector appropriately identifies 26% of AI-written textual content as “possible AI-written.” 

Whereas OpenAI clearly has much more work to do to refine its device, there’s a restrict to simply how good it could actually make it. We’re extraordinarily unlikely to ever get a device that may spot AI-generated textual content with 100% certainty. It’s actually laborious to detect AI-generated textual content as a result of the entire level of AI language fashions is to generate fluent and human-seeming textual content, and the mannequin is mimicking textual content created by people, says Muhammad Abdul-Mageed, a professor who oversees analysis in natural-language processing and machine studying on the College of British Columbia

We’re in an arms race to construct detection strategies that may match the newest, strongest fashions, Abdul-Mageed provides. New AI language fashions are extra highly effective and higher at producing much more fluent language, which shortly makes our current detection device package outdated. 

OpenAI constructed its detector by creating a complete new AI language mannequin akin to ChatGPT that’s particularly skilled to detect outputs from fashions like itself. Though particulars are sparse, the corporate apparently skilled the mannequin with examples of AI-generated textual content and examples of human-generated textual content, after which requested it to identify the AI-generated textual content. We requested for extra data, however OpenAI didn’t reply. 

Final month, I wrote about one other methodology for detecting textual content generated by an AI: watermarks. These act as a kind of secret sign in AI-produced textual content that enables laptop packages to detect it as such. 

Researchers on the College of Maryland have developed a neat method of making use of watermarks to textual content generated by AI language fashions, and so they have made it freely accessible. These watermarks would enable us to inform with virtually full certainty when AI-generated textual content has been used. 

The difficulty is that this methodology requires AI corporations to embed watermarking of their chatbots proper from the beginning. OpenAI is creating these programs however has but to roll them out in any of its merchandise. Why the delay? One motive is perhaps that it’s not at all times fascinating to have AI-generated textual content watermarked. 

One of many most promising methods ChatGPT may very well be built-in into merchandise is as a device to assist folks write emails or as an enhanced spell-checker in a phrase processor. That’s not precisely dishonest. However watermarking all AI-generated textual content would mechanically flag these outputs and will result in wrongful accusations.

Leave a Comment