Comment on Apparently Palantir can access the content of social media accounts that were deleted a decade ago.

• CovfefeKills@lemmy.world ⁨4⁩ ⁨days⁩ ago

  So you acknowledge that the data exists, what you are scared of is being able to search it? Spooky stuffs.

  source

  • mriormro@lemmy.zip ⁨4⁩ ⁨days⁩ ago

    So you acknowledge that bullets exist, what you are scared of is being able to continuously fire them at an extremely high rpm? Spooky stuff.

    You fucking knuckle dragger.

    source

    • CovfefeKills@lemmy.world ⁨4⁩ ⁨days⁩ ago

      You never considered that bullets could be fired at a high rate until an article you saw on lemmy told you to be scared of it?

      source

      • mriormro@lemmy.zip ⁨3⁩ ⁨days⁩ ago

        Jesus, it’s like you’ve melded with the idiot bus.

        source

        • -> View More Comments
  • korazail@lemmy.myserv.one ⁨4⁩ ⁨days⁩ ago

    I’m going to say that this is actually spooky.

    Not that it’s unreasonable, but that the scale of what AI can surveil is so vast that there’s no more personal security-via-obscurity.

    It used to be that unless someone had a reason to start looking at you, anything you did online or off was effectively impossible to search. You might be caught on some store’s CCTV, Or your cell provider might have location pings, but that wasn’t online for anyone and needed a warrant to have the police use it to track your activities. Now cities are using Flock and similar tools to enable tracking vehicles across the country without any reason, and stores are using cloud-service AI cameras to attempt to track your mood as you move through the store. These tools can and have been abused.

    Now, due to the harvesting of this data for AI, anything that’s ever been recorded (video footage, social media posts, etc) and used as training data can be correlated much more easily, long after it occurred, and without needing to be law enforcement with a warrant.

    I’d call that spooky.

    source

    • CovfefeKills@lemmy.world ⁨4⁩ ⁨days⁩ ago

      So you think private and opensource intelligence spontaneously came into existence in the last 5 years because of AI?

      source

      • korazail@lemmy.myserv.one ⁨4⁩ ⁨days⁩ ago

        No, that’s not what I said. Widespread data collection and searching used to be something only state actors could accomplish and there were at least theoretically guard rails. Now the barrier of entry has been seriously reduced, the data is owned by a corporation, and being fed to AI. That has a chilling effect as well as being ripe for abuse.

        I don’t see an upside.

        source

        • -> View More Comments
  • Trainguyrom@reddthat.com ⁨3⁩ ⁨days⁩ ago

    So you acknowledge that the data exists, what you are scared of is being able to search it?

    Usually that’s the insurmountable mountain. Data collection is easy. Formatting, storing and querying the data so you can actually get useful information out of it in a time efficient manner is the extremely hard part.

    For a real world example, the organization I work at does quarterly audits of all of the field offices to make sure all of the field offices are in compliance, checking required document retention, gear, etc. and when an audit finds a requirement that is out of compliance they’re given a task with a deadline to complete said task to bring them back into compliance, and these tasks have visibility all the way up the chain of command to where even the C-levels are reviewing them regularly. I’ve been working a project recently to flag repeated failures of the same audit requirement for the same location and it’s highlighting that some field offices are not actually coming into compliance once these high visibility assigned tasks are completed which when I presented it to leadership it was a revelation just how many field offices are continuously out of compliance.

    Point is, this data is being actively collected and formatted for easy access and there’s still glaring issues being missed due to the difficulty of finding these trends buried in the hundreds of pages of data being generated each quarter per field office

    source

    • CovfefeKills@lemmy.world ⁨3⁩ ⁨days⁩ ago

      Yea for sure more accountable reliable systems would be better than worse systems great point.

      source

      • Trainguyrom@reddthat.com ⁨2⁩ ⁨days⁩ ago

        Uhhh I think you completely missed what I was saying. I was explaining that collecting data is easy, but actually making use of that data is really hard, and gave a real world example of a trend that should be obvious being buried in a mountain of data because there’s simply too much data to sift through

        source

        • -> View More Comments