AI-powered coding is heavy on the brain
Last week in Wonna's Slack, we discussed why working with AI assistants appears differently in ŌURA ring stats compared to autonomous, craftsmanship-driven solo coding: time spent solo coding is recorded as restorative time.
When a developer works with an AI assistant, much of the work shifts from writing code toward reviewing code generated within the boundaries of their prompts and constraints. Reviewing a colleague's code builds on shared history, trust, and context. By contrast, today's AI assistants—still in rapid development—are highly variable from day to day. Trusting them is difficult, which makes the review of AI-generated code particularly draining.
So, how do coding and code review differ as tasks?
When humans write code, the central processes are generative and creative cognitive activities:
Problem framing: Holding the goal in mind—what should the program do?
Working memory and planning: Tracking variables, structures, and program state
Synthesis: Combining knowledge of syntax, libraries, and algorithms
Prospective thinking: Anticipating the effects of the following line of code
Creative problem-solving: Often, there's no single solution, so flexibility is needed
The process of programming is closer to writing text: bringing forth something that does not yet exist.
Code review, on the other hand, is an analytical and evaluative process:
Focused attention: Searching for errors, unclear structures, or improvements
Comparison to models: Reflecting the code against best practices, team standards, and experience
Critical reasoning: Spotting inconsistencies, illogic, or edge-case violations
Metacognition: Asking "Do I understand this? Could someone else understand this?"
Retrospective thinking: Looking over what has already been created—unless proposing changes
The process of reviewing code is closer to proofreading and editing text: critically examining an existing product.
Taking care of the beautiful minds
Employers should approach this shift in software development as both a significant opportunity for increased productivity and joy of releasing great solutions to relevant problems more often, but also as a significant human risk. Experienced, versatile developers with architectural thinking—those who truly know what good software looks like—are at high risk of strain. These are precisely the professionals already in short supply. Many of them also draw recovery and deep-flow satisfaction from focused and independent, craftsmanship-driven coding.
To summarise:
Coding is a creative production process, heavily engaging working memory, ideation, and forward thinking.
Code review is an analytical evaluation process, emphasizing attention, critical reasoning, and drawing on memory (e.g., past bugs, team guidelines).
Code review, especially of AI-generated code, may be more straining to developers, and this should be considered to ensure a sustainable pace in development organizations.
Author
Laura Snellman-Junna
Wonna’s Co-Founder, Laura, helps software teams harness AI-powered development to deliver faster and smarter. With 20+ years in software and consulting, she blends deep technical expertise with a mission close to her heart: making work joyful and developers happier. For Laura, exceptional velocity comes not just from tools and processes, but from teams that thrive together.