AI text Gen The current trend in generating text using large language text bot is growing in popularity and has become an issue along ethical and societal line. These has raised the concern for differentiating between AI generated text and that of Human written text. With this in mind, the OpenAI has set out to solve this issue by launching a tool for detecting text generated using services like the one they have created called ChatGPT. The use of AI generated text models have raised many ethical and societal questions such as – Can AI text generating Bots output misleading information – Could students use this technology to cheat during Assignments, term paper, or even on thesis writing. – Will AI be given credit where excerpts are used for journal articles or paper? – How will this AI bots be cited when giving credit on a journal articles or paper? Recently, it was reported that GPT-3 is able to generate “influential” text that has the potential to radicalize people into far-right extremist ideologies. These was reported by a team of researchers from the the Middlebury Institute of International Studies’ Center on Terrorism, Extremism, and Counterterrorism, (See paper – pdf) However, AI generated text is being embrace and welcome by all especially technology driven youths all across the globe and some of this AI generated text tool used include ChatGPT, DALL-E and a few others under development. Also tech giants such as Google, Meta, Amazon, Microsofts and Apple are investing heavily into AI large language generating models. OpenAI who founded ChatGPT intended it to serve and benefit humanity, but with the raising concern, they have decided that they have a responsibility to ensure its creations have safeguarding in place to prevent or at least minimize their use for harmful purposes. This has lead OpenAI to lunch a new tool this week to enable the detection and differentiation of AI generated text and that of Human, though they warn that it is currently unreliable and 100% effective, as the tool can only identifies 26% of AI-written text as “likely AI-written”. The tool falsely classifies human-written text around 9% of the time. In a statement by OpenAI, they clearly outline the impossibility of 100% uncovering of AI written text “While it is impossible to reliably detect all AI-written text, we believe good classifiers can inform mitigations for false claims that AI-generated text was written by a human: for example, running automated misinformation campaigns, using AI tools for academic dishonesty, and positioning an AI chatbot as a human” and also warn that the tool should not be used as a primary decision-making tool, but instead as a complement to other methods of determining the source of a piece of text.” Other limitations of the new openAI tool for the detection of AI generated text was also spelt out as follow “very unreliable” on short texts under 1,000 characters and it’s “significantly worse” when applied to non-English texts” But OpenAI encourages anyone who is yet to test their new tool called “DetectGPT” to test the OpenAI’s work-in-progress tool for detecting AI-generated text via “DetectGPT” More so, DetectGPT is being created by Stanford University researchers and the demo is available here



By 4niso

Leave a Reply

Your email address will not be published. Required fields are marked *