OpenAI announces “a classifier,” “to distinguish between text written by a human and a text written by AIs from a variety of providers”
The creator of panic de jour, ChatGPT announces its product “correctly identifies” 26 per cent of AI-written text. The more text, the better the system gets.
OpenAI nominates “using AI tools for academic dishonesty” as one of the uses for classifiers.
But as for text-bots eating jobs, perhaps not yet.
OpenAI’s statement is attributed to four authors and 13 contributors.