Comment on Developer creates endless Wikipedia feed to fight algorithm addiction

• Septimaeus@infosec.pub ⁨2⁩ ⁨months⁩ ago

  True, but outside CS the word has come to refer to a certain brand of complex heuristics or ML inference.

  source
• fox2263@lemmy.world ⁨2⁩ ⁨months⁩ ago

  An algorithm usually involves lots of complex calculations and weights. Picking a number from a pool of numbers at random is as simple as it gets.

  source

  • b_n@sh.itjust.works ⁨2⁩ ⁨months⁩ ago

    In comsci, there are no real random numbers. They are all seeded psuedo-random number algorithms. (Unless you integrate with some third party random as a service setup)

    source

    • fox2263@lemmy.world ⁨2⁩ ⁨months⁩ ago

      Yes but the common interpretation of “the algorithm” is that of the social media and YouTube style one. Recommending items of interest etc but easily manipulated by bad actors.

      Wiki random is about as opposite to that as possible.

      source

      • b_n@sh.itjust.works ⁨2⁩ ⁨months⁩ ago

        The modern interpretation, sure.

        And agreed, “random, you might like this” is not as random as “here is a page on red food colouring”

        source
    • weker01@sh.itjust.works ⁨2⁩ ⁨months⁩ ago

      That’s a common misconception. You can measure a lot of ambient noise and extract entropy. Like time between inputs or how long it took an HDD to seek.

      Most modern PC CPUs even have dedicated hardware for generating random numbers from electrical ambient noise. I don’t trust them however.

      source

      • b_n@sh.itjust.works ⁨2⁩ ⁨months⁩ ago

        That’s why I said seeded. Seeded from noise. The random number generator (function) is still an algorithm…

        source
• blind3rdeye@lemm.ee ⁨2⁩ ⁨months⁩ ago

  Right, but in the context of social media feeds, “algorithm” always refers to an algorithm for personalised content.

  source