While there has been a lot of hullabaloo over ChatGPT from OpenAI, the latter has now launched a brand new tool for detecting AI-generated text. The launch has been made official today on the OpenAI blog. ChatGPT is again in the news after having cleared exams for medical and business management schools. Where it seems scary with AI showing capacity to take jobs from the human plate, this AI detector seems like a boon.
The storm called AI-Text
ChatGPT is a free tool that can respond to user prompts of any kind, from fine poetry to management essays—there seems no cap on what the tool cannot give info on.
- It debuted in November last year but has been popular for a lot of wrong reasons too. The latter context refers to issues such as plagiarism and copyright.
- This new AI classifier is trained on model as per dataset pairs based on human-written as well as AI- text on a topic. This aims to demarcate between both variants using a number of providers for handling issues like automated misinformation or academic dishonesty.
- OpenAI has launched the new tool in a beta version for public use and acknowledges the same to not be trustworthy for text below 1,000 characters. A huge challenge that stares the maker is the ability for AI- text to get edited for tricking the classifier.
- The official words goes, “We recognize that identifying AI-written text has been an important point of discussion among educators, and equally important is recognizing the limits and impacts of AI-generated text classifiers in the classroom.”
The repercussions
ChatGPT has already been banned in NY City school districts and a number of third party AI-detectors has also seen popularity among educators to handle such challenges. Google is also testing its own AI-backed chatbot called Apprentice Bard, another AI-powered chatbot. This leverages the LaMDA module of Google. This could be a ChatGPT rival in the times to come.
Summing up
CNBC reports that Google is finally maxing its technology that can generate proper and interactive replies. The report also states, “Another product unit has been testing a new search desktop design that could be used in a question-and-answer form.”
