Empty workstations highlight the growing void of knowledgeable engineers as companies turn to generative AI.
*Recent outages across tech and infrastructure expose a fragile reliance on a dwindling pool of experts who actually care. The scramble to replace them with generative AI threatens the stability of systems that keep economies moving.*
A cascade of high‑profile outages has rattled the tech world. Microsoft 365 stalled, air traffic control hiccuped, United Airlines’ reservation engine faltered, the NSA’s network showed glitches, major banks reported issues, and Apple’s services experienced delays—all within days. The author, a former Stripe engineer, points to these failures as evidence that complex systems survive only because of a handful of smart people who give a shit. At Stripe, he saw ACH—a legacy, flawed, yet generally reliable money‑movement system—kept afloat by experts who patch, bypass, and manually intervene. He argues that the current wave of layoffs and AI‑first policies is stripping away those very custodians, leaving the infrastructure exposed and on the brink of collapse.
In the past week, Microsoft 365, Air Traffic Control, United Airlines' systems, the NSA's network, many major banks and Apple’s systems all experienced disruptions. The author stresses these incidents are not linked to cyber‑attacks or generative AI glitches. Instead, a common thread runs through each: the absence of smart people who give a shit. Those who know the intricate details of each platform are the ones who patch, route around, or manually intervene when the underlying architecture falters. Without that human vigilance, even well‑designed systems tumble, exposing how much modern infrastructure leans on a few dedicated experts.
Generative AI is marketed as a cheaper, tireless substitute for human expertise. It never takes sick days, asks for benefits, or pushes back on bad ideas. The author argues that leadership confuses raw output—words, code, pixels—with the deeper insight that only caring engineers provide. AI lacks the ability to notice edge cases, feel when something is off‑track, or manually resolve unique failures. As companies replace staff with LLMs, layoffs increase and the remaining workforce is forced to babysit prompts instead of mastering the systems they once owned. The result is a hollowed‑out talent pool incapable of sustaining complex operations.
ACH, the backbone of everyday money movement, is described as old, riddled with flaws, and burdened by legacy baggage. Yet it usually works because knowledgeable people at Stripe and elsewhere constantly fix, patch, and manually handle its quirks. The author uses ACH to illustrate a broader principle: any complex system—whether banking, cloud services, or air traffic—relies on individuals with intimate knowledge and a drive to keep it running. Remove those people, and the system’s inherent weaknesses surface, leading to the kind of outages now seen across multiple sectors.
Capitalism begrudgingly accommodates smart, caring workers because they are essential to building and maintaining profitable infrastructure. Yet the same system now views those workers as replaceable by AI, promising higher output at lower cost. The author notes that leadership sees a relief in offloading human labor to machines, ignoring the intangible value of care and ownership. This paradox drives layoffs, erodes expertise, and pushes remaining staff further from the systems they once mastered. The resulting knowledge gap accelerates the risk of future collapses.
The article lists failures in Microsoft 365, Air Traffic Control, United Airlines' systems, the NSA's network, many major banks, and Apple's systems.
The author argues that each disrupted system relies on knowledgeable individuals who patch, route around, or manually handle issues, and their absence leaves the systems vulnerable.
The author claims generative AI is cheaper and tireless but lacks care, edge‑case handling, and the drive to fix problems, leading companies to replace skilled workers and hollow out expertise.
The pattern is clear: without the dedicated experts who understand and care for the inner workings of critical infrastructure, even the most robust systems crumble. As AI replaces human judgment, the safety net of manual insight disappears, and the risk of systemic failure grows. The world must recognize that cheap, tireless machines cannot substitute for the human drive that keeps the global economy humming.
Sources: Hacker News (https://violetpixel.com/blog/2026-11-17-we-need-smart-people-who-give-a-shit/)