OpenAI's AI-generated Text Detector
Created on January 29|Last edited on February 3
Comment
Along with their new premium subscription service, partnership with Microsoft, and rapid growth (and from what I heard, a mobile app for ChatGPT), OpenAI is building a model to detect AI-generated text!
It seems like, in response to a lot of community cautiousness surrounding the potential coherent and perhaps occasionally deceptive gibberish ChatGPT is capable of spewing, they plan to integrate this system with ChatGPT to minimize potential misuse. This is interesting.
A few blog posts back I mentioned how existing applications will need to scale alongside ChatGPT and it looks like OpenAI took the initiative to develop their own system! It's like a 2-for-1 deal. If you use ChatGPT, then you might need this AI-generated text detector.
Apparently, this classifier is trained to detect whether or not a piece of text is AI-generated, not just from ChatGPT but any AI model. it's interesting to think: while we grow more and more integrated with AI, how will we differentiate AI-generated content from what's real? Is relying on another AI to counter that the best option?
References
- Wiggers, Kyle. “OpenAI Releases Tool to Detect AI-Generated Text, Including from Chatgpt.” TechCrunch, 31 Jan. 2023
- Person, and Krystal Hu. “CHATGPT Owner Launches 'Imperfect' Tool to Detect AI-Generated Text.” Reuters, Thomson Reuters, 1 Feb. 2023
- Kelly, Samantha Murphy. “Chatgpt Creator Launches Subscription Service for Viral Ai Chatbot | CNN Business.” CNN, Cable News Network, 2 Feb. 2023,
Add a comment
Tags: ML News
Iterate on AI agents and models faster. Try Weights & Biases today.