Multiple senior developers report that AI-assisted coding replaces the deep-focus satisfaction of solving hard problems the slow way with a lower, more anxious baseline of constant context-switching. James Long (jlongster) described missing "a deep focus and satisfaction... debugging something crazy for 3 days and fully solving it was incredible, and you came away with very deep learnings." Doing that over 20 years built a body of hard-won expertise. He says "all of that is gone" now, and describes swinging between high and low points while feeling stressed every week. Dillon Mulroy corroborated independently: "it's way harder to get joy and satisfaction out of building with ai than it was before," citing "constantly straddling burn out, being far less immersed in hard problems, constant context switching," and uncertainty "of where this is going and how to do it well." The shared complaint is not that AI tools make coding harder. It is that they remove the multi-day immersion in a single hard problem that used to produce both satisfaction and durable learning, replacing it with faster but shallower, more fragmented work. **Cross-Domain Connections**: [[Developer Joy as Leading Indicator]] frames DX as a leading indicator of adoption; this note is the inverse case, where a tool's productivity gains coexist with a measurable DX loss. Also connects to [[Hold the Hard Part Constant to Test the Big Swing]] since both concern what happens when hard problems get engineered away rather than solved. ## Source - [[there's a deep focus and satisfaction that I miss]] — James Long, 2026-07-06 — https://x.com/jlongster/status/2073899675648708967/?s=12&rw_tt_thread=True