The West Forgot How to Make Things. Now It's Forgetting How to Code
This provocative piece argues that the West's diminishing manufacturing capabilities, exposed by defense industry struggles, is a chilling precursor to a looming crisis in software engineering. It posits that a relentless pursuit of short-term optimization and over-reliance on AI are eroding fundamental skills and tacit knowledge, creating a future where critical expertise simply disappears. The article resonated deeply on Hacker News, sparking intense debate about the long-term consequences of current tech trends on talent, corporate strategy, and national resilience.
The Lowdown
The article draws a stark parallel between the West's lost manufacturing prowess and a potential future crisis in software development, attributing both to a relentless focus on short-term optimization at the expense of long-term skill development and institutional knowledge. Author Denis Stetskov, drawing from his experience in Ukraine during the ongoing conflict, highlights how the defense industry's inability to ramp up production of essential items like Stinger missiles and artillery shells stems from decades of optimizing away crucial capabilities and the human expertise to build them.
- Defense Industry Debacles: Examples like Raytheon struggling to restart Stinger missile production by bringing back septuagenarian engineers, Europe failing to meet artillery shell commitments due to shuttered factories, and the multi-year, multi-million-dollar struggle to reproduce Fogbank (a nuclear warhead material) due to lost expertise and documentation, serve as cautionary tales.
- The Fogbank Phenomenon: The Fogbank case is particularly illustrative: a material whose production process was lost, requiring reverse engineering and ultimately revealing that a critical 'impurity' known only by retired workers was essential to its function. This represents the irreplaceable nature of tacit, human-held knowledge.
- Software's Slippery Slope: Stetskov argues that the software industry is making the same mistakes. Companies are reducing junior hiring, relying heavily on AI tools, and inadvertently fostering 'AI-mediated competence' where developers can prompt AI but lack the foundational understanding to debug errors or build complex systems from the ground up.
- Talent Pipeline Collapse: Surveys indicate declining computer science enrollments and engineering leaders anticipating AI will reduce junior hires, suggesting a looming shortage of senior engineers who possess end-to-end system understanding and institutional knowledge.
- The Human Element: The piece emphasizes that money cannot compress the time needed to develop deep engineering expertise (5-10 years for senior roles), and AI, currently, cannot replace the judgment and critical thinking required to navigate complex challenges.
The author concludes with a somber warning: just as the defense industry bet on perpetual peace and optimized itself into fragility, the software industry is betting on AI to solve its talent crisis. If that bet fails, the consequences, like the current struggles in defense, will be severe and enduring, as irreplaceable knowledge simply vanishes with retiring generations.
The Gossip
Optimization's Ominous Outcomes
Many commenters resonated with the article's core premise: the true culprit isn't AI itself, but a pervasive management pattern of short-term cost-cutting. This approach eliminates organizational slack and junior hiring, thus stopping the transfer of tacit knowledge from experienced engineers. Commenters argue that documentation and automation cannot replace field experience or judgment, leading to a loss of essential expertise and a 'hollowing out' of capabilities, similar to the General Electric example cited. This leaves organizations fragile and unprepared for crises.
Coding Competence Catastrophe
The discussion heavily focused on AI's impact on developer skills. Many expressed concerns that AI code generators, described as 'trolls' that produce confidently plausible but partly wrong content, hinder the development of deep understanding in junior engineers. This 'AI-mediated competence' means new developers can't identify AI's errors, leading to a potential 'talent pipeline collapse' and a de-skilling problem akin to the 'way of COBOL.' However, some counter that they are 'flying personally' with AI and wouldn't go back, suggesting not all developers are equally affected by this workflow.
Critiques, Contexts, and Controversies
While many agreed with the article's sentiment, some commenters challenged its framing or specific examples. Critics suggested the piece uses 'decline-porn' and portrays 'the West' as a monolithic entity for clickbait. Others argued that forgetting old ways (like COBOL) is a natural part of progress and that the comparison to manufacturing might be flawed, as coding lacks the same physical cost barriers. There was also a minor tangent questioning the geopolitical implications of the author's Ukrainian background when discussing financial aid.