New AI classifier for indicating AI-written text by OpenAI ✅✅✅

[Sartre]

https://openai.com/blog/new-ai-classifier-for-indicating-ai-written-text/

We’ve trained a classifier to distinguish between text written by a human and text written by AIs from a variety of providers. While it is impossible to reliably detect all AI-written text, we believe good classifiers can inform mitigations for false claims that AI-generated text was written by a human: for example, running https://openai.com/blog/forecasting-misuse/, using AI tools for academic dishonesty, and positioning an AI chatbot as a human.

Our classifier is not fully reliable. In our evaluations on a “challenge set” of English texts, our classifier correctly identifies 26% of AI-written text (true positives) as “likely AI-written,” while incorrectly labeling human-written text as AI-written 9% of the time (false positives). Our classifier’s reliability typically improves as the length of the input text increases. Compared to our https://github.com/openai/gpt-2-output-dataset/tree/master/detector, this new classifier is significantly more reliable on text from more recent AI systems.

We’re making this classifier publicly available to get feedback on whether imperfect tools like this one are useful. Our work on the detection of AI-generated text will continue, and we hope to share improved methods in the future.

 

[HenryObi]

https://openai.com/blog/new-ai-classifier-for-indicating-ai-written-text/

Such product would actually hurt the usage of their tool. If AI content becomes more identifiable, less people would be interested in using chatGPT and similar tools.

 

[zigzagzig]

https://openai.com/blog/new-ai-classifier-for-indicating-ai-written-text/

Tested my AI generated article.

The classifier considers the text to be unlikely AI-generated.

This tool is a joke.

 

[710fla]

Our classifier is not fully reliable. In our evaluations on a “challenge set” of English texts, our classifier correctly identifies 26% of AI-written text (true positives) as “likely AI-written,” while incorrectly labeling human-written text as AI-written 9% of the time (false positives). Our classifier’s reliability typically improves as the length of the input text increases. Compared to our previously released classifier, this new classifier is significantly more reliable on text from more recent AI systems.

Makes sense with all the false positives with other public tools that detect AI

 

[Sartre]

Tested my AI generated article.

The classifier considers the text to be unlikely AI-generated.

This tool is a joke.

amazing

 

[SirLouen]

https://openai.com/blog/new-ai-classifier-for-indicating-ai-written-text/

Paranoia has officially started
Heads will roll

 

[leeVanCleef]

*puts in davinci 003 generated text*

"The classifier considers the text to be very unlikely AI-generated."

 

[Setox]

*puts in davinci 003 generated text*

"The classifier considers the text to be very unlikely AI-generated."

So in-house materials are bulletproof ?! good job Microsoft.

 

[Ababba2]

lol I copy pasted a pretty agrressive answer I found on a thread on Reddit and it says it's AI generated. I mean, I have no proof since it wasn't mine but I'm pretty sure that was human generated.
Also because AI doesn't let you write aggressive stuff

 

[JasonS]

This tool is a complete joke, just copied the AI-generated content into it with ZERO editing.
And it says "The classifier considers the text to be very unlikely AI-generated."

 

[LemonAcid]

Probably not gonna make one reliable content detector in few years

 

[GreenFog]

Such product would actually hurt the usage of their tool. If AI content becomes more identifiable, less people would be interested in using chatGPT and similar tools.

Depends on what you use it for.
Most people don't use it to make easy $ through affiliate marketing, seo spamming, etc