Where are the vibecoded Photoshops?
- The Core Argument: The author challenges the narrative that AI allows unskilled users to prompt and immediately ship complex, professional-grade software. They point out that after years of widespread access to advanced models, the world is not drowning in "vibecoded" equivalents of Photoshop, Excel, or operating systems.
- The "Vibecoding" Accusation: Calling someone’s project "vibecoded slop" has become a destructive social weapon and gatekeeping mechanism. It is used to dismiss AI-assisted work, costing the target immense time and morale to defend while costing the accuser nothing.
- Hypocrisy of the Critics: The accusation itself acts like unverified "vibecoded" content. It is a fast-shipped emotional reaction put out as a factual finding, devoid of definitions, testing, or evidence.
- The Three Levels of Software Work:
- Level 1 (Typing): Mechanical coding, syntax, loops, and memorizing syntax. AI has successfully lowered the barrier to and cost of this layer.
- Level 2 (Verifying): Flow, testing, data structure choices, debugging, and quality control.
- Level 3 (Deciding): Architecture, macro decisions, trade-offs, and long-term design that survives the real world.
- Source of Backlash: The gatekeeping stems from Level 1 programmers who tied their professional identity and self-worth to the physical act of typing code. Because AI made Level 1 cheap, they feel personally threatened and lash out at AI-assisted creators.
- Call to Action: Despite having a rigorous engineering and demoscene background that would allow them to "punch down," the author refuses to weaponize the term. They urge creators to transparently ship their AI-assisted work without apology, and encourage the community to judge projects by their testing and architectural choices.
Hacker News Discussion
- Shift Toward Long-Tail, Bespoke Tooling: Multiple users argue the premise is slightly off because AI isn't meant to build a mass-market "Photoshop replacement." Instead, it is empowering people to build bespoke, narrow-scoped, one-off tools (e.g., custom data scripts, household apps, or personalized pedometers) that solve exact personal needs without needing to learn full-stack development.
- The 3D Printer Analogy: A prominent debate compares vibe-coding to the 2010s hype of household 3D printers. Critics argue that just as 3D printing stalled because CAD design is harder than the actual printing, vibe-coding will stall because software architecture and data persistence are harder than generating basic code. Proponents counter that unlike 3D printing, AI software has zero upfront hardware costs, relies on devices people already own, and lowers the barrier further by translating plain English into functional instructions.
- Moving Goalposts vs. Generative Slop: Some developers express frustration that AI advocates are shifting goalposts from "AI will replace all software engineers" to "AI will build minor scripts." They emphasize that software design remains the difficult part of engineering, and raise concerns over the normalization of low-quality, AI-generated "slop" across tech and art.
- Accessibility vs. Professional Engineering: Commenters note that Level 1 coding was always the easy part, which is why experienced engineers command a premium for architectural foresight. However, making Level 1 universally accessible means a broader demographic of non-techies (the "Uncle Bobs" of the world) can finally build functional tools for themselves and their communities without relying on professional developers.