No Skill. No Taste
This blog post asserts that the rise of LLMs has flooded platforms like Hacker News with low-skill, low-taste applications, creating a deluge of
The Lowdown
The author, a seasoned developer, laments a perceived decline in the quality of applications and projects shared online, particularly on Hacker News. He attributes this to the advent of large language models (LLMs), which he believes create an "illusion of a lower barrier to entry," enabling many to build and share projects that lack both skill and "taste."The core argument is that while LLMs might simplify the technical aspects of creation, they amplify the need for discerning taste. Many new applications are described as "vibe-coded," poorly crafted, and derivative, contributing to noise rather than innovation.The author emphasizes that "taste"--the ability to discern what resonates with an audience or solves a genuine problem in an elegant way--is the true barrier to entry. He cites examples like a simple website that self-destructs without user interaction, which gained popularity purely through its unique concept and "taste," despite technical simplicity. He clarifies that the "sin" isn't using LLMs for development, but rather lacking the skill and taste to produce genuinely valuable work. Ultimately, the author predicts that this influx of low-taste projects will eventually taper off as creators either learn better etiquette or face disappointment, comparing the current LLM hype to the transient "anyone can get rich" narrative of the crypto boom.
The Gossip
Defining Discretion: What is 'Taste'?
Commenters grappled with the elusive definition of "taste." While some found it too vague to be useful, others offered analogies (like Remy in Ratatouille or "code smell") or attempted to pinpoint its essence. The author himself clarified that "taste" might not be globally definable but is discernible within specific niches. There was also discussion on whether taste is subjective, innate, or something that can be acquired and how it might evolve with changing times and cultures. The consensus leaned towards it being a learned, albeit difficult to define, quality essential for good judgment.
AI's Artistic Allegory: Skill, Slop, or Stepping Stone?
A significant portion of the discussion centered on how AI impacts creativity and the quality of output. Some, like `roywiggins`, argued that generative AI can degrade user's taste by habituating them to lower standards, coining the "slot machine effect." Conversely, others celebrated AI's potential to empower new creators: `mlapeter` shared how his 7-year-old uses Claude to build games, fostering learning and self-expression, and `devinprater`, a blind user, detailed how AI has dramatically enhanced his coding productivity and accessibility. This highlights a fundamental tension: is AI a tool that enables "slop" from the tasteless, or a powerful stepping stone for broader creativity and accessibility?
Personal Passion vs. Public Presentation
Commenters debated the distinction between building projects for personal satisfaction and presenting them to a public audience like Hacker News. `lubesGordi` championed the idea of creating for personal joy, regardless of external validation, and not complaining about a crowded market. Others argued for respecting the audience's time, suggesting creators should differentiate their projects in crowded spaces and consider whether their offering truly adds value. The author chimed in, clarifying that building for oneself is positive, but sharing publicly implies a responsibility to consider why others would find it interesting and act in good faith.