In the AI-assisted coding era, developers need to focus less on rote coding skills and more on higher-level abilities like architectural decision-making, code quality management, advanced testing strategies, and effective communication. Curiosity, confidence, and strong soft skills are increasingly vital, as developers must navigate faster feedback loops and act as bridges between technology and business needs.
In this video from the Modern Software Engineering channel, Trisha Gee and Kent Beck discuss the evolving skill set required for developers in the age of AI-assisted coding. They begin by acknowledging the shift in daily activities for developers: with AI tools (referred to as “the genie”) handling much of the code generation, developers now spend more time reading, reviewing, planning, and managing projects rather than typing code. While some traditional skills, like memorizing IDE shortcuts, have become less relevant, the ability to make architectural decisions, recognize when to refactor, and balance short-term versus long-term needs has become even more critical.
The conversation highlights that many of the most valuable developer skills have always been less about syntax and more about judgment, design taste, and knowing when to ask clarifying questions or push back on business requirements. These skills are not rendered obsolete by AI; in fact, they become more important as developers must frequently make higher-level decisions and manage the increased pace and complexity that AI tools introduce. The speakers emphasize that the cost of code creation has dropped, but the friction in the development process can increase if not managed well, making awareness of bottlenecks and the ability to maintain code quality essential.
Testing remains a central topic, with both speakers agreeing that skills developed through practices like test-driven development (TDD) are still highly relevant. While AI can generate tests and suggest coverage improvements, developers must still understand what needs to be tested, interpret results, and ensure the system is robust. The ability to use advanced testing techniques—such as mutation testing and fuzz testing—becomes even more valuable, as AI can automate much of the grunt work, allowing developers to focus on higher-level test strategy and analysis.
Curiosity and confidence are identified as crucial skills for developers in this new landscape. The speakers note that while AI can act as an infinitely patient tutor, developers must cultivate the confidence to learn new tools and the curiosity to explore unfamiliar concepts. The ability to articulate high-level goals, break down problems, and ask the right questions—both of AI tools and of oneself—enables developers to get the most out of augmented coding environments. This self-awareness and willingness to learn are seen as highly leveraged skills that will only grow in importance.
Finally, the discussion turns to soft skills, which are often undervalued in technical hiring and self-assessment. Communication, prioritization, negotiation, and the ability to translate ambiguous business requirements into actionable technical tasks are all highlighted as essential. As developers increasingly act as a bridge between technology and end users, these interpersonal and intrapersonal skills become even more critical. The fast feedback loops enabled by AI mean developers must exercise these skills more frequently, making them a key differentiator in the age of AI-augmented software development.