A sharp divide is growing inside tech companies: executives can’t stop evangelizing AI, while the individual contributors (ICs) actually building products remain deeply skeptical. A widely discussed analysis on Hacker News, scoring 176 points, offers a compelling framework for why this gap exists: it comes down to something fundamental about how work itself is structured at different levels of an organization.
The core argument is elegantly simple. Executives have always managed non-deterministic systems. People call in sick, projects slip without warning, features get built in ways that technically meet objectives but miss the point entirely. Managing chaos is the job description. AI, with all its quirks, fits right into that worldview.
Why Executives See a Dream Tool
From a management perspective, AI actually behaves better than many human systems. As the Hacker News analysis points out, LLMs have several properties that make them attractive to anyone used to managing unpredictable teams:
- They produce output regardless of time of day or task difficulty
- Their failure modes are well-defined and increasingly understood (hallucinations, context limitations)
- The capability envelope is getting mapped out quickly, unlike individual humans whose strengths and weaknesses take months to uncover
For an executive who’s spent years adding determinism to inherently chaotic organizations by using processes, levels, ladders, and SOPs, AI looks like the most predictable “employee” they’ve ever managed. That’s a powerful draw.
Why ICs See a Liability
Individual contributors live in a different reality. Their value comes from being reliably precise: writing correct code, nailing the analysis, producing designs that hold up under scrutiny. The more deterministic your output, the better you are at your job.
AI introduces non-determinism into exactly that space. And the skepticism isn’t irrational. Three reasons stand out:
- It’s not as good as they are. A highly trained human focused on a specific task will often beat an LLM, especially for long-running work that requires connecting multiple systems or deep domain intuition. The overhead of fixing AI mistakes can genuinely cost more than doing the work yourself.
- It changes the nature of the job. You go from doing the work to managing something that does the work. The skills that got you hired (deep focus, precision, domain knowledge) aren’t necessarily the skills that make you good at supervising an AI.
- It’s tied to self-worth. When executives talk about AI making everyone more productive, ICs can hear that as: the things you’ve spent years mastering are about to matter less. Whether or not that’s what’s being said, it’s a reasonable thing to feel.
Speed vs. Quality: The Hidden Variable
One of the most practical observations from the analysis: organizations that bias toward speed over quality see significantly more IC adoption of AI. Engineers at startups are using AI tools to move faster, even if the output isn’t necessarily higher quality. Organizations that prioritize quality often see the opposite pattern.
This tracks with what’s happening across the industry. AI coding assistants are gaining traction fastest at companies where shipping quickly matters more than shipping perfectly. Where precision is the currency (infrastructure, security, financial systems), adoption remains slower and more contested.
What This Means for AI Strategy
This framing has real implications for how companies should approach AI adoption. Mandates from the top, like “everyone must use AI,” land differently depending on where you sit. For executives, it’s obvious. For ICs, it can feel like being told your expertise matters less.
Companies getting this right are doing a few things differently:
- Letting adoption be task-specific rather than blanket mandates
- Acknowledging the quality tradeoff honestly instead of pretending AI makes everything better
- Redefining what “good” looks like when AI is in the loop, rather than applying old metrics to a new workflow
The friction between executives and ICs on AI isn’t about one group being right and the other being wrong. It’s about two fundamentally different relationships with uncertainty. Until companies acknowledge that gap instead of steamrolling through it, the Slack debates and comment thread arguments will keep going.
The full discussion is worth reading for anyone navigating AI adoption decisions at their organization.