Home 🏠

📱 Social Media Algorithm Simulator

Machine Learning-Based Content Recommendation

⚠️
Educational Simulation: This demo shows how social media algorithms track your behavior (not just your clicks!) to predict what you'll engage with. Watch the right panel to see what signals the algorithm detects and how it adapts to your behavior in real-time.
Posts Seen
0
Content Diversity
100%
Time Spent
0:00

📱 Your Feed

Session: 0:00
📱

Ready to Start!

Click "Start Feed" to begin scrolling. The algorithm will learn from your behavior.

🧠 What the Algorithm Sees

📊 Current Post Signals

Scroll to a post to see signals...

🎯 Learned Preferences

Interact with posts to build your profile...

🔮 Next Posts Predicted

Algorithm will predict content you'll engage with...

🧠 How Social Media Algorithms Really Work

Unlike the movie recommender (which uses your explicit ratings), social media algorithms track implicit behavioral signals:

Signals Tracked:
• Dwell Time: How long you look at a post (even without interacting)
• Mouse Movement: Did your cursor hover over the post?
• Scroll Speed: Did you slow down or skip past quickly?
• Likes & Clicks: Explicit engagement (worth more points)
• Skips: Actively skipping content sends a negative signal

The Algorithm's Goal: Maximize engagement (time spent, interactions), NOT your satisfaction or wellbeing. This creates a "filter bubble" where you see more of what keeps you scrolling, even if it's not what you consciously want.

💡 Try This: Notice how spending just 10 seconds looking at a post (even without liking it) teaches the algorithm that you want more similar content. Your behavior speaks louder than your intentions!

⚠️ Watch For: The "rabbit hole" effect - engaging with one controversial or emotional post can quickly shift your entire feed toward more extreme content, because those posts generate high engagement.