Tech work was never the clean, rational craft outsiders imagine; greed, not AI, is dismantling engineering apprenticeship and the institutional knowledge it produced.
Key Takeaways
The junior engineer pipeline was effectively killed in 2024: output was optimized for, apprenticeship abolished, and the seniors who would have emerged from it will not exist in five years.
“Sara” is the archetype of what was lost: inherited institutional knowledge (a 3am cron job running since 2016, a USB stick from a dead mentor) that no agent can replicate or replace.
CEOs returning from offsites with agent demos promised 30% engineering cuts by Q2; the people who signed the lists knew the DORA metrics and Goodhart’s Law already predicted the collapse.
The greed argument is explicit: same capital logic that offshored factories and exploits cobalt mines, rebranded as AI transformation.
When Sara and her equivalents retire, there is no replacement pipeline; the machine that made her was already dismantled before anyone noticed she existed.
Hacker News Comment Review
Consensus that AI is a proximate cause but not root cause; the real driver is investor and board pressure using AI demos as cover for headcount cuts that were already wanted.
Sharp split on whether systems will collapse soon and force a hiring rebound, versus a more pessimistic view that the apprenticeship gap is permanent and no rebound fixes it.
Several readers flagged the irony of debating AI co-authorship in the comments of an essay arguing for human craft: at least one commenter suspected Claude-editing mid-read, which others pushed back on as itself a signal of degraded baseline expectations.
Notable Comments
@arian_: “now we have to debug code we didn’t write, can’t fully understand, and definitely can’t explain in a code review” – concrete operational cost of agent-generated codebases.
@pdp: predicts AI-driven code debt will cause system collapses that trigger a developer hiring spike larger than current supply can fill.