Algorithms


Brenna’s Digital Detox post about algorithms got me thinking about where algorithms and opaque magic bits of code intermediate what I see online. It’s definitely less than it has been, since unplugging from Facebook and reducing my Google exposure. But, still. These are the ones I’m aware of… (I’ll update this as I think of stuff and/or realize something’s managed by The Algorithm™)

Algorithmic stuff

  • Amazon - searches and listings are based on what they think I’ll buy. apparently, it’s a lot of chinese knock-off electronics for some reason. who knows how it determines what crap I should buy, or what crap I’m likely to buy even if I shouldn’t, but if I see it’s ON SALE NOW I might be 4% more likely to buy it on an impulse and have it delivered by 9am tomorrow…
  • Apple Mail. Junk filter - it’s pretty accurate, but I have no visibility into how it does what it does.
  • Apple Music. “Hey siri, play my radio station” - I have no idea how it comes up with the playlist. It’s almost always perfection. Based on likes/dislikes and previous plays? some input based on what friends are listening to? some trending stuff? opacity.
  • Apple News. I don’t know how it decides which stories I need to see. It’s not very accurate, though.
  • DuckDuckGo search results. I assume it’s purely relevance-related, but who knows? And how is “relevance” defined? This is 99% of my search engine use.
  • Google search results. Only in the rare cases where DDG doesn’t find something I need. Who knows how this crap is sorted/filtered/promoted.
  • Office365 Mail Focused/Other inboxes (which messages are deemed “important” and which are “noise” and can be dealt with later)
  • New York Times’ “For You” section.
  • Reddit front page & Popular - based roughly on what subreddits I subscribe to, with some wackiness involving activity somehow. Definitely not just based on chronological order. Who knows…
  • Twitter feed. It sure ain’t reverse-chronological order. I have no idea why it shows me what it shows me.

Non-algorithmic stuff

OK. It’s impossible for something to be completely non-algorithmic. Sorting is an algorithm. Pagination is an algorithm. But, things that don’t obviously intermediate in the content or people I see…

  • D2L Brightspace - courses, course content, etc. is all straightforward. There is an adaptive learning thing available, but we haven’t licensed it.
  • Mastodon.social feed. I think it’s mostly reverse-chronologial order, but maybe it’s not.
  • NetNewsWire RSS feeds - it’s strictly reverse-chronological-order, with no cleverness applied. Perfect.

See Also

comments powered by Disqus