Was this AI trained on an unbalanced data set? (Only black folks?) Or has it only been used to identify photos of black people? I have so many questions: some technical, some on media sensationalism

    • nieceandtows@programming.dev
      link
      fedilink
      arrow-up
      1
      ·
      1 year ago

      Again, I’m not disagreeing on systemic racism in the police at all. That is a big issue that needs to be solved. Just saying that this doesn’t have to be related to it, because the technology itself has some issues like this. The vinegar is in the food, yes, but lemon is naturally sour. Even if there is no vinegar, it’s gonna be sour. Attributing everything to vinegar wouldn’t make food better. It would just make it difficult to identify issues with individual ingredients.

    • Gareth Kitchen@stroud.social
      link
      fedilink
      arrow-up
      1
      ·
      1 year ago

      @fishidwardrobe As far as the UK is concerned (re facial recognition) I recall the latest study has found false positives disproportionately higher for Black people and statistically significant.

      The UK Police thought this acceptable and have continued the roll out of this tech. A judgement call that bakes a little bit more systemic racism into UK Policing with little to no accountability.

      https://science.police.uk/site/assets/files/3396/frt-equitability-study_mar2023.pdf

      PS. I’m not academically qualified to comment on the paper, but take an interest in these things.

      @nieceandtows