back to article OpenAI offers error-prone AI detector amid fears of a machine-stuffed future

OpenAI has released a free online tool designed to predict whether a passage of text was generated by AI or written by a human. Dubbed the "AI Text Classifier", the software is powered by a language model and rates the likelihood a chunk of text was generated by an AI model on a five-point scale that goes from "very unlikely" …

  1. ecofeco Silver badge

    Let the AI war begin!

    Well that escalated quickly.

  2. Sora2566

    Very emotionally charged

    I get the distinct impression that the author of this piece has strong opinions on AIs being used in this way...

    1. diodesign (Written by Reg staff) Silver badge

      Hand bitten

      Articles are generally a collective effort, unless it's marked opinion or comment, in which case, it's solely the author's view.

      As a whole, we're done with this charade. There are so many examples like this and this.

      Congrats, Redmond, you bought yourself a multi-billion-dollar moron.


      1. Dimmer Bronze badge

        Re: Hand bitten

        I will throw this out foe fun:

        There are articles today about asking it to write something nice about Trump and Biden.

        Refused write anything about Trump.

        Do AI fear being canceled?

  3. Pascal Monett Silver badge

    Rated unclear

    Now, now, those Vultures should be commended for having helped the vast domain of statistical computing get off its feet.

  4. T. F. M. Reader Silver badge

    Business model

    Can't decide if this is OpenAI going back to its not-for-profit roots or switching to ransomware, a.k.a. "we'll flood you with BS and will provide you with an AI-pow(d)ered nose for a few coins", model.

    1. CatWithChainsaw

      Re: Business model

      Well obviously it's the latter.

  5. amanfromMars 1 Silver badge

    Playing Catch Up in Diabolical Fields where Mares and Stallions have Bolted to Heavenly Pastures

    What very few folk anywhere realise is El Reg bots/AI are instrumental in pioneering leading machine learning. A situation they have been publishing freely for quite some time now unhindered by unwarranted and wilful negative intrusion and viral venal infection.

    1. amanfromMars 1 Silver badge

      Re: Playing Catch Up in Diabolical Fields where Mares and Stallions have Bolted to Heavenly Pastures

      Well? Yes? No? .... Hello? Is there anybody out there?

  6. Johnb89

    Just remember what you spat out?

    Wouldn't the easier-but-not-AI-therefore-not-sexy way to detect AI generated text be to remember what text had been spat out by it, and let teachers etc check submissions against that? Not for single sentences, but faked essays would be a doddle.

    Too obvious?

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like