Big Tech alum says the individual contributor role is 'probably over' — AI is making everyone a manager

In the evolving landscape of the tech industry, a significant shift is occurring with the advent of artificial intelligence in the workplace. Former OpenAI employee Philip Su, who left to establish his startup Superphonic, has raised eyebrows with his assertion that the traditional role of individual contributors (ICs) is fading away. Suis, who has also held positions at Microsoft and Meta, recently expressed on Substack that AI has fundamentally altered the responsibilities of software engineers. He claims that rather than focusing solely on building products, engineers are increasingly tasked with managerial duties like setting priorities, resolving conflicts, and providing feedback. He argues that this shift is not due to AI surpassing human coding abilities, but rather because maximizing productivity now requires engineers to engage in activities traditionally associated with management. On a recent episode of the podcast "A Life Engineered," Su reiterated that the IC role is likely coming to an end, as engineers are turning to AI for assistance in their work. This raises an intriguing question: if everyone is managing AI, do we still need human managers? Su believes the answer is yes, citing a persistent "human coordination problem" that AI cannot address. This theory aligns with the current trend of organizational flattening within major tech companies, where fewer managers are overseeing larger numbers of ICs. Notably, companies like Tesla and Amazon have been streamlining their management layers to enhance the IC-to-manager ratio. Furthermore, tech leaders are embracing a "founder mode," a concept that contrasts with what Paul Graham terms "manager mode." Interestingly, there may be a resurgence of managerial roles—albeit in a new form. Jeanne DeWitt Grosser, COO of Vercel, suggests the emergence of the "agent manager" position, designed to oversee AI systems. Tools that facilitate this management are also advancing. Su has noted that he now reviews AI-generated code with less scrutiny than before, likening it to managing a team of highly capable individuals rather than inexperienced interns. As the industry continues to adapt to these changes, the balance between human oversight and AI assistance will remain a topic of discussion and innovation.