OpenAI’s new tool can help you recognize text written by ChatGPT.

It has OpenAI. Device released To help you determine whether text is written by humans or AI. However, the creator of chatgpt warns that it is equivalent Blade RunnerThe Voight-Kampf test can also go wrong.

The tool includes a box where you can paste text that is at least 1,000 characters long. It is then followed by something like “The classifier considers the text highly unlikely to be AI-generated” or “The classifier considers the text to be AI-generated.

I tested chatgpty by asking him to write an essay on the state of bird migration, which the diagnostic tool described as “probably AI-generated.” Meanwhile, it rates several human-authored articles as “highly unlikely to be AI-generated.” Although the tool can throw up false flags in both directions, my (small sample size) test shows at least some accuracy. Still, OpenAI cautions against using the tool alone to determine content authenticity. It also works best with text of 1,000 words or more.

The startup faced pressure from educators after the November release of its ChatGPT tool, which said AI-written content could sometimes pass for human text. A natural-language model can generate essays in seconds based on simple text questions – even for graduate business and law exams – giving students an exciting new opportunity to cheat. As a result, New York Public Schools banned the bot from their WiFi networks and school devices.

While ChatGPT’s arrival has been a hot topic of late, and even extends to media outlets looking to automate SEO-friendly articles, the bot is a big deal for OpenAI. The company reportedly received a $10 billion investment from Microsoft earlier this month, which it plans to integrate into Bing and Office 365. OpenAI is said to have discussed selling its stock for $29 billion late last year. .

Although ChatGPT is currently the best natural language AI model available publicly, Google, Baidu and others are working on competitors. It’s plausible that a former researcher at Google’s LAMD quit the search giant last year after claiming the chatbot was a scam. (The human tendency to outsource emotion and consciousness to algorithms is a concept we’ll be hearing a lot about in the coming years.) It’s possible that Google released very strict versions of its chatbot in beta because of ethical concerns. When the genie is out of the bottle, it will be interesting to see how long the suspension lasts.