The Fear of ‘Good Enough’ AI and the Death of Software Craft

The author worries that AI will stabilize at producing competent but uninspired software that most people accept. Because incentives reward speed and cloning, AI agents will flood the world with slop while suppressing artisan craft and unique design. If user apathy persists, the craft of software may die—and no one will care.

Key Points

• AI may stall at a ‘90% good enough’ level that satisfies markets while eroding standards of software craft.
• AI tooling gravitates toward median, template-like solutions, making unique, high-touch design and behavior harder to achieve.
• Slop predates AI and stems from incentives: big companies clone and kill artisan work; AI will accelerate this dynamic.
• The dream of end-user programming may be overestimated; widespread tech apathy could keep most users passive consumers.
• The author’s core fear: software craftsmanship could die—and most people wouldn’t care enough to notice.

Sentiment

The community is notably divided. A large contingent sympathizes with the article's concern about quality erosion but frames it as an old problem merely accelerated by AI rather than created by it. The prevailing mood is one of weary resignation — many acknowledge the problem but doubt anything can or will change. A smaller but vocal group pushes back harder, arguing the fears are fundamentally overblown or that market forces will self-correct. The discussion frequently drifts from software craft specifically into broader anxieties about capitalism, wealth inequality, and societal collapse, suggesting these concerns resonate beyond the article's narrow focus on software.

In Agreement

• AI will accelerate the commodification of software by making it trivially cheap to produce 'good enough' output that nobody bothers to refine
• The Pareto principle is now automated: AI delivers 80% solutions instantly, and the missing 20% of craft, care, and originality will simply be abandoned
• Real job displacement is already happening across translation, graphic design, call centers, and bookkeeping, with no clear replacement jobs materializing
• The wealth gap will widen as AI enriches capital owners while displacing workers who have few options to 'move upmarket'
• AI-generated output converges toward a homogeneous median, making distinctive and innovative software increasingly rare
• Society may face instability if large portions of the workforce are displaced without adequate economic restructuring

Opposed

• This fear is nothing new — capitalism has always driven toward minimum viable quality, and consumers have always rationally chosen cheap over good, long before AI
• Product quality has actually improved dramatically in many categories (TVs, phones, medical technology, battery life) while prices have fallen
• Every major technology transition — from printing to industrialization to the internet — initially degraded quality but eventually stabilized at a new equilibrium with quality persisting in niches
• AI job displacement claims are exaggerated, with companies using 'leveraging AI' as cover for already-planned layoffs and overhiring corrections
• LLMs show diminishing returns requiring exponentially more resources for linear improvements, making the feared disruption less likely to materialize at scale
• The article mourns craft without proposing solutions, which some argue makes it an unactionable lament about market dynamics that cannot and should not be changed