Why Executives Love AI While ICs Remain Skeptical

The AI adoption debate within companies reveals a striking divide: executives enthusiastically embrace AI while individual contributors remain deeply skeptical. This tension plays out in Hacker News threads, internal Slack debates about coding agents, and company-wide AI mandates. The root cause? Executives and ICs operate in fundamentally different worlds when it comes to determinism and control.

Executives: Comfortable with Chaos

Executives have always managed non-deterministic systems. Their careers involve navigating unpredictable human behavior, shifting market conditions, and complex organizational dynamics. Consider the daily chaos they handle:

• Team members calling in sick unexpectedly
• Projects derailing without warning
• People reacting unpredictably to announcements
• Features built that technically work but don't align with product vision

This mirrors chaos theory in mathematics—when individual agents act with different inputs and utility functions, nonlinear systems emerge. A manager's core job becomes modeling this chaos and aligning utility functions across the organization.

AI fits naturally into this worldview. Large language models exhibit characteristics of well-behaved chaotic systems: they produce consistent output patterns despite unpredictable specific results. Executives appreciate that:

• LLMs work reliably regardless of time or task difficulty
• Failure modes are well-defined (hallucinations, context limitations)
• Capability boundaries are becoming clearer
• The system is more predictable than human teams

After years of adding structure through processes, levels, and standard operating procedures, executives see AI as another tool to introduce determinism into inherently chaotic systems.

ICs: Precision in a Deterministic World

Individual contributors operate differently. Most ICs focus on specific problems with clear inputs and outputs. Their value comes from being reliably precise—writing correct code, producing accurate analysis, creating designs that withstand scrutiny.

While ICs certainly face non-determinism (unclear requirements, flaky systems, shifting priorities), they're evaluated on deterministic output. Quality and speed become the primary metrics, with the weight depending on organizational culture.

This creates friction when AI enters the picture:

It's not as good at specialized tasks. A highly trained human expert often outperforms AI, especially for complex, multi-system tasks requiring deep domain intuition. The overhead of fixing AI mistakes can exceed doing the work manually.

It changes the nature of their work. ICs shift from doing work to managing something that does work. The skills that made them valuable—deep focus, precision, domain knowledge—don't necessarily translate to being good at AI oversight.

It threatens professional identity. When executives talk about AI making everyone more productive, ICs hear that years of skill development might become less valuable. Even if unintended, this perception is reasonable.

The Quality vs. Speed Divide

Organizational culture significantly impacts AI adoption. Companies prioritizing speed over quality see more IC enthusiasm for AI tools. Startup engineers in my network actively use AI to accelerate tasks, even if quality doesn't improve.

Organizations valuing quality often see resistance. AI doesn't inherently improve quality—it can sometimes make specific tasks worse. ICs trained for precision in their domains find AI's non-deterministic output unsettling.

Where the Friction Comes From

The AI perception gap stems from fundamentally different work experiences. Executives manage non-deterministic systems and have built careers around it. ICs operate in more deterministic worlds and are evaluated on delivering precise, reliable output.

This framing explains why the same AI tool looks completely different depending on your role. What executives see as productivity enhancement, ICs might view as a threat to their professional identity and work quality.

The divide isn't about AI's technical merits—it's about how different organizational layers experience work itself. Until companies acknowledge this fundamental mismatch, AI adoption will continue creating friction rather than harmony.