Glue work is still harmful in the age of AI (www.seangoedecke.com)

Recent commentary highlights that while Large Language Models (LLMs) continue to revolutionize coding practices, the concept of "glue work"—the often unrewarded, non-technical tasks that facilitate team cohesion and project success—remains prevalent but is diminishing in significance. The author outlines several reasons for this shift: AI reduces the need for extensive coordination between team members, leading to a scenario where individuals can handle projects that once required multiple collaborators. Additionally, fewer junior hires mean less need for oversight and onboarding, while the automation of report generation through LLMs eliminates much of the tedious paperwork previously managed by dedicated "glue" individuals. However, the piece cautions against the assumption that AI can completely replace the interpersonal aspects of glue work, emphasizing that genuine communication among engineers is irreplaceable. While LLMs excel at creating written content, nuanced interactions and detailed project updates require personal engagement to maintain healthy professional relationships. As engineers adapt to these changes, the focus should remain on coding and leveraging AI tools effectively, while still being open to performing necessary glue work when it contributes to project outcomes. This commentary serves as a reminder for the AI/ML community to rethink their approach to collaboration and recognize the evolving landscape of team dynamics amidst technological advancements.