AI giant OpenAI has halted the development of its AI detection software, citing low accuracy rates. The artificial intelligence classifier, which was launched on January 31, was designed to help users like teachers and professors distinguish between human-written text and AI-generated text However, the tool was shut down on July 20 due to its low accuracy.
In the blog post announcing the tool's launch, OpenAI stated that the AI classifiers are no longer available because of their poor performance. The company acknowledged that the current tool had limitations and was "very inaccurate" in validating text with fewer than 1,000 characters . It also had issues flagging human-written text as AI-generated.
OpenAI is now working on finding more effective ways to identify AI-generated content. The company is incorporating feedback and researching improved text proven techniques. It also plans to develop and deploy mechanisms that will enable users to determine whether audio or visual cont ent is AI- generated.
This is not the first product from OpenAI to come under review. On July 18, researchers from Stanford University and the University of California, Berkeley, published research showing that OpenAI's flagship product, ChatGPT, has deteriorated significantly with age. The ability of ChatGPT-4 to accurately identify prime numbers dropped from 97.6 percent to 2.4 percent over the past few months. Additionally, both ChatGPT-3.5 and ChatGPT-4 exhibited decreased ability to generate new lines of code.
OpenAI's decision to halt the development of its AI detection software and its ongoing efforts to improve its products demonstrate the company's commitment to refining and enhancing its AI technology. The company is dedicated to addressing the limitations and challenges associated with AI -generated content to ensure more accurate and reliable results in the future.
