The Evolving Role of Software Engineers: Productivity, Management, and the AI Paradox
The world of software engineering is undergoing a profound transformation, with AI taking center stage as a game-changer. The idea that every IC engineer is now a front-line manager is gaining traction, but is this shift truly beneficial for productivity? The answer, it seems, is complex and multifaceted.
The Rise of the AI Manager
Cameron Etezadi, CTO of LaunchDarkly and former VP of engineering at IBM, makes a compelling argument. He believes that the traditional IC role is obsolete, and engineers now require project planning and cross-team coordination skills, essentially becoming managers. This perspective aligns with Gartner's prediction that smaller software engineering teams will dominate by 2029, with tiny teams of just two to three engineers. If this trend continues, the software engineering role might evolve into a managerial position, but will it enhance productivity?
Productivity Metrics: More Code, Faster, or Meaningful Outcomes?
The question of productivity is complex. Daniel Wang, CTO at Citizen Health, challenges the notion that AI coding tools directly translate to productivity. He emphasizes that software exists to solve problems, and if AI enables the shipment of 10 times more code without improving customer outcomes, it's not truly productive. This perspective highlights the need to shift focus from code output to decision quality and outcomes. Metrics such as cycle time, rollback rate, and system reliability become crucial, along with qualitative questions about the right solutions and customer problem-solving.
However, Ameya Kanitkar, founder and CTO of Larridin, argues that companies are still tracking irrelevant metrics like lines of code and velocity points, which don't accurately gauge value. This approach could steer engineering teams away from true productivity, as they might chase volume without real value. The pressure to produce more quickly is evident, with some engineers even using AI agents to work overnight, creating a constant background pressure that is draining.
The Paradox of AI-Driven Exhaustion
The paradox is that while AI tools might make engineers feel productive, they often lead to exhaustion. David Holz, founder of Midjourney, observes that engineers are feeling both extremely productive and drained. This disconnect between feeling productive and being productive is a significant challenge. Most companies lack reliable ways to track AI's impact on outcomes, making it difficult to assess true productivity gains.
Kanitkar suggests that the new managerial aspect of their work is causing fatigue, as engineers constantly context-switch to manage AI agents. As teams shrink, this constant agent-babysitting could become the new norm, but its productivity benefits remain uncertain.
The Future of Engineering: A Balancing Act
The evolving role of software engineers presents a delicate balance. While AI tools can enhance productivity, they also introduce new challenges. The key lies in finding the right metrics and strategies to measure and manage this new era of engineering, ensuring that productivity is not just about code output but about meaningful customer outcomes.
