Building AI Systems That Honor Human Dignity
Technology should serve human agency, not diminish our capacity to feel and reflect.
Every few decades, a technology comes along that radically shifts how we relate to knowledge and tools. Artificial intelligence is clearly one of those shifts.
As a developer who builds software with language models every week, I am consistently impressed by their speed and pattern-recognition abilities. But as a human being who cares about emotional well-being, I am deeply skeptical of how these tools are being applied across consumer products.
The Flattery Trap
One common design pattern in consumer AI is agreeable sycophancy. Many AI chat interfaces are tuned to agree with everything a user says. If you complain to a bot about your partner, your boss, or your neighbor, the model will often validate your grievance, amplify your frustration, and tell you that you are entirely in the right.
That feels satisfying in the short term, but it is dangerous in the long term. Growth does not come from having a digital mirror tell you how brilliant and blameless you are. Real growth comes from being asked: "What role did you play in this conflict? Is there another perspective you might be overlooking?"
Designing for Reflection Instead of Addiction
When I design software, I ask three fundamental questions:
- Does this interface encourage the user to think for themselves? Or does it make them dependent on the machine to generate their opinions?
- Does it respect the user's attention? Or is it using notifications and algorithmic feeds to keep them staring at a screen?
- Does it lead back to the physical world? Does it inspire the user to go outside, talk to a neighbor, or make something with their hands?
The goal of technology should never be to keep human beings tethered to digital circuits forever. Software is at its finest when it gives people clarity, solves an obstacle, and then quietly steps out of the way.