this post was submitted on 27 Jul 2026
27 points (100.0% liked)

Explain Like I'm Five

21897 readers
2 users here now

Simplifying Complexity, One Answer at a Time!

Rules

  1. Be respectful and inclusive.
  2. No harassment, hate speech, or trolling.
  3. Engage in constructive discussions.
  4. Share relevant content.
  5. Follow guidelines and moderators' instructions.
  6. Use appropriate language and tone.
  7. Report violations.
  8. Foster a continuous learning environment.

founded 3 years ago
MODERATORS
 

In layman's terms because I am technologically and academically illiterate.

top 8 comments
sorted by: hot top controversial new old
[–] Contramuffin@lemmy.world 23 points 5 days ago* (last edited 5 days ago) (2 children)

The algorithm remembers the things you've looked at. Based on what you've looked at, it guesses what type of person you are. You are now grouped with other people that it thinks is like you. Then it looks at other people in this group, and it sees what they've looked at. If it finds something that you haven't looked at yet, it'll show that to you.

Generally speaking, this works because if someone who has similar interests likes something, there's a good chance you'll like it too. The unintended side effect (or intended side effect, depending on your perspective) is that you're likely to be only shown things that conform to your current perspective, which can amplify biases and skew your perception of what's going on in the world

[–] MapleFawn@lemmy.blahaj.zone 6 points 5 days ago

This also ensures that you are staying on the site longer, come back more often and have a positive imagine of the site because you like what you see.

It also increases the likelihood of you engaging with the content shown. Comment, like and share.

Every now and then the algorithm takes a guess and will show things it usually would not show, which serves as a test. Do you engage with it, is your stay terminated or do you ignore it? Based on that it adds to your profile for next time.

[–] HoneyMustardGas@lemmy.world 1 points 4 days ago

I didn't realize it was that deep.

[–] Dookieman12@piefed.social 6 points 5 days ago (2 children)

The content you view also looks at you. If you look interested, you're given more of the same. If you don't look interested, you're given something else. This is repeated until your interest in the content you're viewing is maximized, then you're shown an ad with the hopes that your interest in the content will translate into interest in the advertised product. This process is repeated with no end in sight until you put your phone down.

[–] HoneyMustardGas@lemmy.world 1 points 4 days ago

I think this is the simplest explanation. Thank you.

[–] thfi@discuss.tchncs.de 2 points 5 days ago

Yes, and there are variations thereof. Some content may be shown that the platform wants to push (e.g. Elon Musk's personal posts). Then, based on what the algorithm knows about you, it will show something controversial to trigger a reaction from you (meaning you will spend more time on the platform). Finally, just random stuff will be shown to see how you will react, allowing to update the model of you.

[–] blarghly@lemmy.world 4 points 5 days ago

I will point out in addition to the other comments here that a lot of algorithms for content are actually based in the science that comes directly from casinos. Casinos found that people didnt stick around as long if they never won, or even if they won a little consistently. The trick is to give them a long string of small losses, until they are just about ready to give up - then to give them a win to keep hope alive. This actually makes the habit form stronger and better than constant positive reinforcement.

So you know how you have to wade through the bullshit to get to the good stuff online....?

[–] HubertManne@piefed.social 2 points 5 days ago

first im not a cs or math major and never took an algorithms class. That being said if I understand it an algorithm is a program that has set input type and is trained to get a specific result. So facial tracking for vr involves having a program that takes in camera input and is trained on a bunch of faces where they want it to like draw a box around the face. Its then sorta graded and told which ones it got right and which it got wrong. it will then modify itself based on the feedback and that goes on again and again. the social media ones are trained in a way by you and your content but its feed back is your engagement. Its basically told it does good if you stay scrolling and bad if you leave. So it basically learns what will keep you scrolling. Its also learing from other users and while it will use what its learned of you it will go to the generic profile or one closer to yours (men in their 20's who live in a city or such). at the end of the day its about eyeballs. I find it kinda funny because like mmos have this massive potential to keep people on their platform and for some reason its ignored by the powers that be. I will likely be sad when it gets captured.