post text

Picture this:

  1. You type on Google “laptop won’t turn on”
  2. Google now knows you have a broken laptop and can estimate how desperate you are to fix it.
  3. Because it knows how desperate you are, it can increase shop prices proportionally.

You are going to pay the maximum they get you to pay.

That’s algorithmic pricing.

The more companies know about you, the more they can predict and sell how desperate you are to other stores out there.

An internet-connected car knows much more about you than you realize. A smart TV also knows what you like. Your Alexa knows if there is a problem in the home.

Privacy is much more than just sensitive data.

It’s about not giving leverage away.

Because algorithms will use it against you.

Be safe out there.

Nostr.

  • solsangraal
    link
    fedilink
    English
    arrow-up
    67
    ·
    2 months ago

    it would seem like someone’s name is the least useful data point

    • rtxn@lemmy.world
      link
      fedilink
      English
      arrow-up
      53
      ·
      2 months ago

      That’s the whole thing about browser fingerprinting too. Take the set of internet users who have a particular version of a particular operating system, a particular version of a particular browser, having a particular set of typefaces installed, having a particular language preference, and you’ll find yourself in the intersection of all of them.

      • Trailblazing Braille Taser@lemmy.dbzer0.com
        link
        fedilink
        English
        arrow-up
        35
        ·
        2 months ago

        Remember, kids, it only takes 32 bits to uniquely identify any person on the planet. That’s 32 yes or no questions. Of course, they have to be perfectly crafted questions, but identifying power of fingerprinting must not be underestimated.

      • WolfLink@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        7
        ·
        edit-2
        2 months ago

        To add to this.

        Here’s a website to help you check your own trackability:

        https://coveryourtracks.eff.org

        It can also help give you advice on how to improve your privacy.

        Things that help: (tldr use adblockers but otherwise it’s really about blending into the crowd)

        • using a browser that respects privacy (e.g. not Chrome)
        • using a “popular” browser (using something weird can help narrow it down to you because not many people are using that)
        • (Firefox is a good browser choice. Safari is fine, Edge is probably ok. Avoid Chrome).
        • using an ad blocker
        • using a VPN can sometimes help but can also sometimes hurt because, again, it helps narrow you down.
        • using a popular device can make it harder to track

        Hard to track: uses Firefox with uBlock origin. Maybe using a popular VPN. Uses an iPhone or a popular model of Android like the Pixel (although Google owning Android/Pixel might mean they get your data anyway…)

        Actually very easy to track: uses a niche Chromium-based browser you got from GitHub with niche GitHub project as blockers and a little/known VPN. Uses a niche brand of smartphone with a niche non-Android based OS on it.

        • GoofSchmoofer@lemmy.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          2 months ago

          That was interesting.

          Is there an add-on that changes the header information from an HTTP request to show bogus but common identifying information?

          That is crowdsource the most common configurations and set that as the default in the add-on, so now you look like just another face in the crowd.

          • WolfLink@sh.itjust.works
            link
            fedilink
            English
            arrow-up
            4
            ·
            2 months ago

            It’s more than just the header information. The graphics and audio checksum can give away details of what your device is, even if those details don’t match what header information you are sending. That mismatch is itself information they can use.

    • XTL@sopuli.xyz
      link
      fedilink
      English
      arrow-up
      2
      ·
      2 months ago

      Names actually have a really high collision rate, so for collecting information they’re not good. You don’t want all the different John Smiths’ data clumped together. They’re useful once you start sending personalized stuff in though.