A former OpenAI, Microsoft, and Meta executive, Philip Su, argues that the traditional role of the individual contributor (IC) — especially in software engineering — is becoming obsolete as artificial intelligence (AI) reshapes how work gets done in tech. In a January Substack post and a recent podcast appearance, Su claimed that engineers are increasingly spending time on tasks once reserved for managers — like setting priorities, resolving conflicts, and guiding output — largely because AI now handles much of the actual coding. Rather than writing software themselves, engineers are acting more like coordinators of AI systems that generate the work.
According to Su, this shift doesn’t happen because AI is inherently better at coding but because effective use of these tools requires skills historically associated with management. Engineers must oversee, tweak, and integrate AI outputs, which blurs the line between individual work and managerial oversight. In this view, the “halcyon days” of dedicated IC roles may be fading as professionals spend more time in roles akin to managing sophisticated AI agents rather than producing code directly.
This argument aligns with a broader industry trend known as the “great flattening,” where companies like Amazon, Meta, and others are reducing traditional management layers in favour of flatter organisational structures. While some companies have historically tried to operate with minimal managers to boost efficiency, Su insists that human leadership remains essential — not to directly produce work, but to guide and coordinate AI systems and teams, solving the inherently human problems that AI can’t.
Looking ahead, Su suggests that new roles — such as “agent managers” who supervise AI agents — will become more common. These positions will require people who can effectively marshal AI tools and integrate them into human workflows, rather than traditional engineers focused on hands-on coding. As AI improves and takes on more of the technical heavy lifting, the balance between individual technical work and managerial responsibilities is likely to continue shifting.
