The world of software engineering is undergoing a profound transformation, and the role of the individual contributor (IC) is being redefined. The rise of AI has not only changed the way engineers work but has also turned them into de facto managers, according to Cameron Etezadi, CTO of LaunchDarkly and former VP of engineering at IBM. This shift in dynamics raises an intriguing question: Are engineers becoming more productive, or are they merely being tasked with new responsibilities?
Etezadi's perspective is supported by research from Gartner, which predicts that 60% of organizations will have smaller software engineering teams by 2029, with some teams potentially consisting of just two to three engineers. This trend suggests that the traditional IC role may be evolving into a management position, where engineers are expected to coordinate and plan projects across teams.
However, the notion of productivity in this new era is complex. While AI coding tools are hailed as the future by Big Tech, their impact on productivity is not universally accepted. Daniel Wang, CTO of Citizen Health and former director of engineering at Uber, argues that productivity should be measured by customer outcomes, not just the volume of code produced.
Wang emphasizes the importance of tracking metrics such as cycle time, rollback rate, escaped defects, and system reliability. He believes that focusing on these factors provides a more accurate assessment of an engineering team's performance. Ameya Kanitkar, founder and CTO of Larridin, shares a similar sentiment, suggesting that companies are often tracking irrelevant metrics like lines of code or velocity points, which do not truly reflect value.
The pressure to produce more code faster can lead to a sense of exhaustion among engineers. Kanitkar observes that some are even using AI agents to work overnight, creating a constant pressure to feed new work to the agents. This raises a deeper question: Are engineers becoming more productive, or are they merely being asked to do more with less?
The new managerial aspect of engineering work, as Kanitkar suggests, may be a significant contributor to fatigue. Engineers are constantly context-switching to manage multiple agents, which can be draining. As engineering teams shrink, this constant agent-babysitting could become the new norm, but its productivity benefits remain uncertain.
In conclusion, the evolution of the IC role in software engineering is an intriguing development. While AI tools may enhance productivity, the definition of productivity itself is being challenged. As the industry continues to adapt, it is essential to strike a balance between embracing new technologies and ensuring that engineers' well-being is not overlooked. The future of software engineering may be more about management and coordination than coding, but the true measure of productivity remains a subject of ongoing debate and exploration.
