Nobody ever gets credit for fixing problems that never happened (2001) [pdf]
Capability Traps & Non‑Monotonic Improvement
- Several commenters map the article’s “capability trap” to software and AI: you can’t improve without temporarily getting worse (lower throughput while refactoring, training, etc.).
- Managers often expect improvement to be monotonic and costless, so any dip in short‑term performance is punished, reinforcing the trap.
- A “trap” is framed as: easy to enter (cut corners under pressure), hard to exit (no slack to invest in capability).
Y2K, Preparedness Paradox, and Probability
- Y2K is a central example: many argue the “nothingburger” outcome was precisely because huge effort went into remediation (old COBOL systems, utilities, banking).
- Others note some countries did little and appeared fine, with hypotheses like imported software, newer systems, or small localized issues being invisible externally.
- Discussion touches on the “preparedness paradox”: success makes prevention look like overreaction, plus people’s binary thinking about risk and misunderstanding of probabilities.
Organizational Incentives: Firefighting vs Prevention
- Strong theme: organizations reward visible heroics (late‑night outages, crisis saves) more than quiet prevention.
- “Struggling” or chaotic departments often get more budget and praise than stable, well‑run ones—sometimes for fixing problems they themselves caused.
- Attempts to track and report proactive work sometimes fail in environments with stack ranking or “exceeds expectations only” cultures; this can incentivize not fixing issues until they explode.
- Some suggest surfacing “near misses” and pain to leadership instead of silently absorbing it, so the system can learn, but this can backfire in underfunded or blame‑heavy orgs.
Simplicity, Elegance, and Rewarding Complexity
- Many note that elegant, simple solutions look obvious in hindsight and are undervalued compared to baroque systems that “look hard.”
- Promotions, publications, and kudos often follow perceived complexity, encouraging over‑engineering and bureaucracy.
- AI tools may be accelerating code and solution complexity, making defensive skepticism more common among engineers.
Management, Technical Insight, and Systemic Bias
- Persistent tension between technical staff and non‑technical management: prevention work is invisible to those far from the system, who may see “nothing broke” as “nothing was done.”
- Some argue technical backgrounds help management understand and value preventative work; others counter that good management is a separate skill and engineers can also be poor managers.
- Several see the issue as fundamentally human and systemic—optics, short attention spans, and cognitive biases—rather than unique to tech.
Prevention in Society & Individual Coping
- Parallels drawn to healthcare, safety engineering, infrastructure, and Covid policy: prevention rarely gets glory, and overreaction vs prudent preparation is hard to distinguish after the fact.
- Commenters describe personal adaptations: refusing to play hero games, strategically gaming incentives, or changing roles/industries to places where prevention and “boring reliability” are better rewarded.