The "brains in jars" mental model treats developers as isolated, interchangeable units, ignoring collaboration and social dynamics. This simplistic view leads to burnout, low productivity, and an inability to solve complex problems because it devalues essential work like mentoring and learning out loud.
Psychologist Cat Hicks created a metric to measure how often developers feel forced to prioritize short-term output over long-term goals. High scores on this "overproduction pressure" metric strongly correlate with self-censorship, reduced mentorship, and other cascading negative effects on team health and code quality.
A large-scale study found individual developer cycle times are highly variable week-to-week. An engineer appearing "fast" one week may be "slow" the next. This suggests performance is driven by environmental friction and problem complexity, not a stable individual trait, making it pointless to categorize people as fast or slow.
Beyond simple code generation, developers are using AI as a personal learning tool. They leverage it to explore legacy codebases, understand unfamiliar programming languages, and build skills. This reframes AI's role from a simple productivity accelerator to a powerful engine for knowledge discovery and upskilling.
Psychologist Cat Hicks' research revealed that many developers, including highly experienced ones, suffer from "code review anxiety." This intense fear of judgment prevents them from engaging in a process critical for learning and quality control, and they often suffer in silence, afraid to speak up about it.
The fear that AI will devalue hard-won skills creates a visceral "skill threat" for developers. This triggers a defensive state that impedes rational problem-solving and adaptation. To counter this, leaders must create psychological safety and "safe off-ramps" that emphasize continuous learning and resilience.
Working with AI agents causes significant cognitive load. The solution isn't just more information (cognitive scaffolding), but "metacognitive scaffolding"—teaching developers strategies for how to think, learn, and manage focus. This approach, which includes deliberate practice, makes them more effective and less exhausted.
Not all developers need to be product-obsessed. Rather than forcing uniform motivation, leaders should establish "superordinate goals"—overarching objectives that allow individuals with diverse personal interests (like mastering a language) to contribute meaningfully toward a shared outcome, respecting their individual drives.