The creator of ChatGPT launches a tool to detect texts written by artificial intelligence

The popularization of systems Artificial intelligence (AI) as ChatGPTcapable of automatically generating all types of texts, are opening a deep social debate on the impact of this emerging technology. What if we can no longer distinguish between what a human writes and what a machine writes? Will that accelerate our mistrust?

That fear grows as the chatbot —capable of answering falsehoods by presenting them as true— is increasingly used by students, academic researchers and even journalists to carry out their work.

That has led to OpenAIthe developer company of ChatGPT, to release a new tool free to predict if a writing has been artificially generated. The company that popularized that AI is now publishing another AI to detect it.

We’re developing a new tool to help distinguish between AI-written and human-written text. We’re releasing an initial version to collect feedback and hope to share improved methods in the future. https://t.co/4dQE3dX6vX

—OpenAI (@OpenAI) January 31, 2023

“Unreliable”

However, this sorter “not totally reliable”. OpenAI itself points out that its initial phase tool correctly identifies only 26% of the texts written by ChatGPT or other similar AIs, while in 9% of cases it labels text generated by humans as artificial. “Our classifier has a number of important limitations. It should not be used as a primary decision-making tool, but as a complement to other methods to determine the origin of a text,” the company explains.

Related news

So why does OpenAI publish a tool that fails? The company has assured that it has shared it “to gather opinions” and to improve its method. The language models like the one used by ChatGPT improve when training with new questions to solve. Thus, relying on the public to experiment with this “imperfect” classifier is the quickest (and cheapest) way to perfect your product. That’s why they did the same with the launch of their chatbot.

sam altmanone of the founders of OpenAI, has warned in the latter about the uncritical use of ChatGPT because it incurs bugs and is “impossible to make it perfect”. Still, they will continue to experiment with the AI. An experimentation that also responds to the will of the company —in which Microsoft has invested $10 billion—to get ahead of the competition and be the first to lead the emerging market for what may be the technology of the decade. In recent days, researchers had posted their solutions to detect AI generated text. Taking its own, OpenAI seeks to continue gaining ground.