The 'People You May Know' suggestion list

That random person you met at a party once isn't "suggested" by fate. You’re just a dot on a giant digital map, and the algorithm is obsessed with closing the loops.

It uses a trick called "triadic closure." If you know Alex, and Alex knows Sam, the system bets you’ll click on Sam’s profile just to see what they’re up to. It’s digital matchmaking for your curiosity.

By scraping your contacts and tracking shared locations, it builds a "shadow profile" of your entire life. It often knows who you might know before you even remember their name.

Wait, how do they build a profile for someone who never signed up?

Think of it as a jigsaw puzzle where you’re the only missing piece. Even if you’ve never clicked 'Sign Up,' your friends have. When they upload their contact lists, the algorithm sees a recurring phone number that doesn’t have an account yet.

By cross-referencing these lists, the system builds a 'ghost' identity. It knows who your inner circle is and what your interests likely are based on the people who already have you in their phone.

You aren't invisible; you're just a 'data-shaped hole' in their network. They’ve already built the cage; they’re just waiting for you to walk inside and claim your name.

So they just assume I'm exactly like the people in my contacts?

It’s a principle called "homophily"—or more bluntly, "guilt by association." The algorithm operates on the cold assumption that humans are unoriginal creatures who cluster in predictable herds.

If five of your closest contacts are obsessed with vegan keto diets and indie folk music, the system marks your "hole" with those same labels. It doesn't need to ask you; it just looks at the company you keep.

To the machine, you aren't a unique soul with a complex inner life. You're simply the statistical average of the ten people you text most often.

Does that mean they're actually snooping on my private conversations?

They don't need to read your 'I'm running late' texts to profile you. That’s the amateur way. They focus on the 'metadata'—the digital envelope, not the letter inside.

Even if your messages are encrypted, the system sees the frequency and duration of the connection. It’s like a spy watching a house; they don't need to hear the conversation to know who’s visiting.

By mapping these patterns, they calculate your 'closeness score.' To the machine, messaging a specific person at 2 AM says more about your lifestyle than any words you actually typed.

How does the machine actually use that 'closeness score' against me?

It’s the blueprint for your social hierarchy. If you have a high score with someone, the algorithm pins their face to the top of your feed and ensures you never miss their updates. It’s not a timeline; it’s a curated VIP list.

This score also predicts when you’re about to get bored. If your interaction with a 'close' friend dips, the system might trigger a 'memory' or a notification to bait you back into the app and restart the habit.

They aren't just tracking your friendships; they’re engineering them. By prioritizing certain people, the machine decides who stays relevant in your life and who fades into the digital background.