What Happens If AI Makes Things Too Easy for Us?
TL;DR
Psychologists at the University of Toronto published a commentary in Communications Psychology (February 2025) arguing that removing too much effort from human tasks via AI may erode learning, motivation, and meaning.
Key Points
- The concept of 'friction' – difficulty, struggle, discomfort – is backed by psychological research as essential for deep understanding and durable memory.
- AI is already automating document summaries, code generation, and even emotional support at accelerating pace.
- The central question: what do humans lose when effort disappears from daily life – and what are the long-term costs?
Nauti's Take
Nobody wants to manufacture difficulty for its own sake. But the research is clear: those who never struggle rarely learn deeply.
That applies to students just as much as to developers who outsource every line of code to an LLM. The insidious thing about comfort is that it feels good – even when it causes long-term harm.
The real design challenge for AI is not maximum frictionlessness, but intelligent calibration: where does automation genuinely help, and where does it quietly kill competence? Ignoring that question produces tools that delight in the short term and create dependency in the long run.
Sources