So yeah, AI is everywhere now. Netflix knows what you want to watch, Spotify curates your vibe, except you still spend 20 minutes browsing Netflix looking at slop and never actually picking anything. The interface feels "effortless"?
I have opinions.
This episode is basically me ranting about the mess of AI personalization. We want smart systems, but we don't want creepy surveillance capitalism selling our data to 1,650 "partners." So how do we design this stuff without being terrible people?
What matters:
- Transparency is non-negotiable, tell users why they're seeing what they're seeing ("based on your recent activity")
- Bias is real, your AI is only as good as the data you train it on, and historical data is packed with discrimination
- GDPR exists for a reason, pre-checked cookie boxes are illegal, and explicit consent means explicit
- Accessible design helps everyone, everyone's disabled at some point (holding groceries, broken arm, whatever)
Real examples: X/Twitter got fined with a 120M Euro fine for their mess of a blue check system. Meanwhile, Zalando and Booking actually designed with GDPR in mind from the start.
The point: Personalization and minimalism aren't opposites. The best design removes barriers, not features. Start with privacy. Don't use dark patterns. Respect your users, and they'll stick around.
Exercise: Pick one app that feels both smart and simple. Use it mindfully. Notice how it works. Let me know what you find.
And yeah, see you probably next year. We are where we are. 🐱
