ChatGPT generates cancer treatment plans that are full of errors, study shows
www.businessinsider.com
Researchers at Brigham and Women's Hospital found that cancer treatment plans generated by OpenAI's revolutionary chatbot were full of errors.

ChatGPT generates cancer treatment plans that are full of errors — Study finds that ChatGPT provided false information when asked to design cancer treatment plans::Researchers at Brigham and Women’s Hospital found that cancer treatment plans generated by OpenAI’s revolutionary chatbot were full of errors.

    • lonke@feddit.nu
      English
      3·
      10 months ago

      ChatGPT is GPT 3.5 & GPT 4, as far as I’m aware.

      3.5 is also very capable when it comes to programming, for any well known framework or language. It’s not as capable, but it is still very capable.

        • lonke@feddit.nu
          English
          1·
          10 months ago

          Well, the data it was trained on had a cutoff point in 2021 which would explain that.

          I’ve used it (GPT 3) a fair amount for Unity, and I’m fairly pleased with the results, it’s saved me a fair amount of time. Implementing object pooling and editor window dialogues for scene translation management for example.

          Of course, programming knowledge is required for it to be of consistent use, which, on second thought, may not be at all obvious.