The Scarcity of Taste in the Age of AI Slop

As AI makes the production of software and content effortless, the traditional technical barriers that once filtered for quality have disappeared. This shift devalues the labor of making but elevates the importance of 'taste,' an intuitive sense of quality forged through experience and friction. Ultimately, the most vital human role in an automated world is curation: the ability to judge what is worth keeping amidst a sea of generated mediocrity.

Key Points

• AI has collapsed the distance between an idea and its artifact, removing the 'production wall' that once acted as a filter for quality.
• Taste is not mere preference but a wordless verdict of quality developed through the friction of failure and manual labor.
• The removal of friction in the creative process prevents new practitioners from developing the judgment necessary to distinguish between 'fine' and 'right.'
• Economic incentives favor the high-speed production of 'slop'—indifferent, functional output—over the slow, invisible work of human curation.
• In a world where making is free, the primary craft shifts from the ability to produce to the ability to decide what deserves to exist.

Sentiment

The overall sentiment is cautiously aligned with the article's concern: many commenters agree that AI raises the value of taste and may weaken skill formation, but the community remains divided over whether this is an existential decline in craft or a shift in where human judgment is applied.

In Agreement

• Taste is developed through repeated mistakes, friction, and consequences, and LLMs risk letting people skip that formative process.
• AI-generated code and prose often converge toward bland average patterns that are functional but lack individuality, context, or deeper judgment.
• Relying on LLMs can weaken architectural thinking, repo-level mental models, and other software craftsmanship skills that require sustained practice.
• The convenience of AI follows the broader pattern of tools reducing hard mental effort, which can create dependency and lower the average level of capability.
• The ability to refuse plausible but mediocre output is becoming a central scarce skill in software and writing.

Opposed

• AI will not eliminate creativity because the people who care about craft and originality will continue to push beyond default model output.
• Most software and writing were already mediocre, so LLMs may reveal or accelerate an existing condition rather than create it.
• LLMs can free people to focus taste on higher-level product, system, hardware, or domain decisions instead of low-level implementation details.
• Lowering the barrier to programming may introduce new contributors with strong ideas and different forms of taste.
• Judging work primarily by whether it seems AI-generated can become an unhelpful shortcut that ignores whether the argument is sound.