Programming for enjoyment does not need a productivity argument

Developer Chris Kiehl argues that programming can remain a satisfying hobby even when AI can produce the same result faster. He describes writing a buggy Breakout clone by hand, exploring stronger Java types and struggling with recursion schemes because understanding the ideas is rewarding. His distinction is between paid work’s economic goals and personal exploration; the latter does not need to prove that it saves time.

Chris Kiehl · Oct 17, 2026

Why it matters

Engineers can protect curiosity and skill-building without making every personal project compete with workplace output metrics.

More context

This short personal essay offers concrete examples of learning without a promised payoff. An editorial next step is to choose one small problem you want to understand and let that learning goal determine how much automation you use.

Keep in mind

This is the author’s perspective, not evidence about employment trends or a controlled study of learning. His claim about what AI could generate is personal context rather than a benchmark.

Original read