this post was submitted on 20 Feb 2026
234 points (97.2% liked)
Technology
81612 readers
4283 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
The Facebook feed is an advanced algorithm that knows a shit ton about what to feed you to keep you engaged. It's not just the cookies from sites you visit. They track what thumbnails get you to stop scrolling. They track the way a human eye moves and how far your thumb glides across the screen.
Point is, if it's all scantily clad thirst traps, thats what gets your attention. If you see one, and you stop to take a screenshot for an article you're writing about how it's all thirst traps, then every third item will be another thirst trap.
Facebook doesn't care if you want to see that content. Their goal is to keep your eyes on Facebook. If it makes you mad enough to comment, that's engagement.
I didn't read the whole article, so maybe the author addresses this, but what you see on Facebook is a funhouse reflection of your own interests.
When I was still using Instagram reels, I was always amazed how quickly the algorithm figured me out. If I hesitated for even a second on a reel, it would amend my next ones immediately. I assume the real trick is comparing it to the average time spent on a reel, everyone spends longer on a wall of text reel, but when I stop on a Linux reel for an extra second, I'm immediately in the 1% for engagement.
I read something years ago about how your phone keyboard tracks your recommended words, it knows if you're more likely to type apple or Apple, or if you type soup more than average, and any app that gets that data and compares it to the baseline has an instant, in depth profile on you.
This is exactly why using something like FUTO Keyboard is essential, nowadays.
I'll have to try that. Can you comment at all on the quality of the autocorrect and ability to learn your vocabulary? I don't want to connect my keyboard to a server farm but I certainly do want it to learn to help me type more efficiently on this tiny keyboard.
Well, if you enable the model fine-tuning, it should be fine. I do not commit a lot of mistakes, and I usually correct manually, though.
This is why I love the idea of Cromite and other “antifingerprinting” efforts, not simply blocking but spoofing and plausibly randomizing as many metrics as they can.
I wish there was some way to distribute that to the masses. Like maybe a crazy hardware zero day, and it’s only used to stealth load anti fingerprinting on as many devices as it can.
Exactly, if I look at my facebook feed, besides posts from people and groups I actually follow, the "suggested" posts are all stuff that matches my interests. gaming, music, movies.
They hadn’t used Facebook in years, so this is the default algorithm state — rage bait and bate-bait.