The Fogbank pattern applies to software: optimize away the human pipeline, enjoy the savings, then discover the knowledge is gone when crisis hits.
Key Takeaways
Defense lost the ability to manufacture Fogbank material by eliminating the human pipeline; software engineering is following the same substitution arc via AI.
Before AI, offshore contract teams were the cheaper substitute; AI is the latest iteration of a longer deskilling cycle, not a new phenomenon.
Money was never the binding constraint when defense capability collapsed; knowledge was, and it cannot be reconstructed on demand.
The article argues the timelines for skill atrophy in defense manufacturing and software engineering are structurally identical.
Hacker News Comment Review
Strong consensus that the root cause is management cutting junior hiring and organizational slack, which halts tacit knowledge transfer that no documentation or automation can replicate.
Multiple commenters flagged that the article is visibly AI-assisted, with choppy prose and LLM-isms throughout, making the deskilling argument self-demonstrating in the text itself.
Debate on whether junior, mid, and senior labels still signal genuine engineering depth or merely familiarity with current tooling stacks and dysfunctional org patterns.
Notable Comments
@Animats: calls AI code generators “trolls” that produce confident, plausible, partly wrong output with no flow, forcing humans into error-hunting with no sense of progress.
@anonzzzies: documents an earlier wave: formal verification and “teaching how to think” curricula were replaced by job-aligned Java courses between the late 1980s and early 2000s, predating AI by decades.
