• alexsantee@infosec.pub
    link
    fedilink
    English
    arrow-up
    64
    arrow-down
    1
    ·
    1 month ago

    It’s a shame to see the journalist trusting an AI chat-bot to verify the trustworthiness of the image instead of asking a specialist. I feel like they should even have an AI detecting specialist in-house since we’re moving to having more generative AI material everywhere

    • ohulancutash@feddit.uk
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 month ago

      Did they though? They mentioned a journalist ran it through a chat bot. They also mention it was verified by a reporter on the ground.

      It’s like criticising a weather report because the reporter looked outside to see if it was raining, when they also consulted the simulation forecasting.

    • Riskable@programming.dev
      link
      fedilink
      English
      arrow-up
      1
      arrow-down
      9
      ·
      1 month ago

      It’s not a shame. Have you tried this? Try it now! It only takes a minute.

      Test a bunch of images against ChatGPT, Gemini, and Claude. Ask it if the image was AI-generated. I think you’ll be surprised.

      Gemini is the current king of that sort of image analysis but the others should do well too.

      What do you think the experts use? LOL! They’re going to run an image through the same exact process that the chatbots would use plus some additional steps if they didn’t find anything obvious on the first pass.