Pluralistic: AI and amateurism (15 Jun 2026)
🔗 **Source:** [Pluralistic: AI and amateurism (15 Jun 2026)](https://pluralistic.net/2026/06/15/vernacular/)
> [!note]+ Document Notes
> tags: [ai-criticism, vibe-coding, user-agency, vernacular-technology, ai-bubble, digital-rights]
> author: Cory Doctorow
> source: Pluralistic
> reference: https://pluralistic.net/2026/06/15/vernacular/
> date created: {{date:YYYY-MM-DD}} {{time:HH:mm}}
>
> # 🏷️ Title: AI and amateurism
>
<!-- deep-read:sections start -->
## 🧭 Table of Contents
- [📝 Summary](#-summary)
- [📖 Main Content](#-main-content)
- [🗒️ Notes](#️-notes)
- [📚 References](#-references)
- [🗣️ Paraphrase](#️-paraphrase)
- [📰 Newsletter](#-newsletter)
- [✅ Action Items](#-action-items)
- [🧪 Hypotheses](#-hypotheses)
- [📊 Methodology](#-methodology)
- [❓ Questions](#-questions)
- [🔮 Future Research](#-future-research)
- [💡 Implications](#-implications)
- [🧠 Learning Styles Assessment](#-learning-styles-assessment)
- [📊 Accessibility Assessment](#-accessibility-assessment)
- [🌍 DEIB Assessment](#-deib-assessment)
- [✅ Logical Validity](#-logical-validity)
- [🛡️ DARVO Analysis](#-darvo-analysis)
- [🎯 Conclusion](#-conclusion)
## 📝 Summary
- Cory Doctorow argues that "vibe coding" (AI-generated personal software) has value when used for individual projects but becomes problematic when applied to production code or commercial software.
- He distinguishes between user-empowering technology (like early internet tools) that gives people direct control versus corporate-controlled platforms that extract value from users.
- The AI industry focuses on replacing programmers to justify its $1.4 trillion investment, but the real potential lies in personal automation tools.
- Unlike early web adoption driven by workers demanding access, AI adoption is being pushed by bosses threatening workers.
- The most promising applications involve individuals creating idiosyncratic tools for their own specific needs, continuing the tradition of "the street finds its own use for things."
## 📖 Main Content
Doctorow explores the paradox of his position as both a technology enthusiast and AI skeptic, rooted in cyberpunk's dual nature: excitement about technology's possibilities alongside wariness of its oppressive potential.
### Historical Context
Doctorow traces his computing journey from a 1970s cardboard CARDIAC computer through early Apple computers, modems, and eventually becoming a digital rights activist. This background shaped his belief in user-controlled technology that enables direct creation rather than just consumption.
### The iPad Problem
He cites the iPad as an example of consumption-focused technology designed to prevent user innovation without corporate permission, representing "cyberpunk as a demand, not a warning."
### Vibe Coding's Promise and Peril
**Personal Use**: Vibe coding extends the power of shell scripting and desktop automation to broader audiences, enabling personal projects like plane-tracking apps that serve individual needs.
**Commercial Misapplication**: Problems arise when vibe-coded prototypes are expected to scale to production systems serving millions, or when inexperienced users submit vulnerable code to open-source projects.
### Industry Dynamics
The AI industry promotes vibe coding as programmer replacement to justify massive investments, contrasting with early web adoption driven by worker demand rather than management mandate.
### Vernacular vs. Corporate AI
Doctorow distinguishes between:
- **Corporate AI**: Centralized, homogeneous, recognizable output that looks "boring and washed"
- **Vernacular AI**: Personal tools running locally, unsuited for production but valuable for individual automation
## 🗒️ Notes
- Cyberpunk philosophy balances technological excitement with awareness of oppression risks
- "Nothing about us without us" principle: people understand their own needs better than technologists working on their behalf
- Meta-tools like Hypercard and Scratch enable non-programmers to create software matching their specific desires
- Early web champions were concerned about both positive potential and dystopian risks, contrary to naive optimism narrative
- The $1.4 trillion AI bubble brings in only tens of billions annually with worsening unit economics
- Local AI models exist but are treated as sideshows compared to expensive "frontier models"
- Personal software is "deplatformed from its inception" and inherently unmonetizable
- Wreccer project by Naomi Novik aims to enable personalized interfaces through APIs in the vibe coding era
## 📚 References
- [The Reverse Centaur's Guide to Life After AI](https://us.macmillan.com/books/9780374621568/thereversecentaursguidetolifeafterai/)
- [CARDIAC cardboard computer](https://www.instructables.com/CARDIAC-CARDboard-Illustrative-Aid-to-Computation-/)
- [Wreccer project](https://github.com/wreccer)
- [Telegraph article on AI money-wasting](https://www.telegraph.co.uk/news/2026/06/04/ai-is-the-greatest-money-wasting-scheme-humanity-has-ever-i/)
## 🗣️ Paraphrase
Doctorow argues that AI's value lies not in replacing professional programmers but in empowering individuals to create personal automation tools. While the AI industry promotes vibe coding as a path to eliminating skilled workers and centralizing control, the real promise resembles early internet culture where "the street finds its own use for things." Personal AI tools running locally can serve individual needs without requiring corporate permission or scaling to production systems. This vernacular use of AI continues the tradition of user-controlled technology that gives people direct power over their digital tools, contrasting with consumption-focused platforms designed to extract value. The key distinction is between technology that empowers users to create versus technology that requires permission from corporations to modify or extend.
## 📰 Newsletter
**AI That Actually Works For You**
Cory Doctorow's latest piece brilliantly captures why "vibe coding" (AI-generated personal scripts) might be the most exciting development in user empowerment since WordPress plugins. While the AI industry burns through trillions trying to replace programmers, the real magic happens when individuals create weird, wonderful tools for their own specific needs—like that journalist's plane-tracking alert system.
Think about it: WordPress democratized web publishing by letting non-developers create websites. Now AI could democratize automation by letting anyone script their own digital workflows. The catch? It only works for personal projects, not production code.
This connects to WordPress's philosophy of giving users control over their digital presence. Just as custom themes and plugins let you shape your site exactly how you want, personal AI tools could automate your unique workflows without corporate gatekeepers.
**Resources:**
- [Wreccer](https://github.com/wreccer) - Building personalized recommendation systems
- [The full article](https://pluralistic.net/2026/06/15/vernacular/) for Doctorow's complete take
## ✅ Action Items
#### DevRel
- Explore AI-powered developer tools that enhance rather than replace human creativity
- Document use cases where vibe coding helps developers prototype ideas quickly
- Create educational content distinguishing between personal automation and production code requirements
- Develop guidelines for evaluating AI-generated code contributions to open source projects
#### OSPO
- Policy & Governance: Establish guidelines for AI-generated code contributions and review processes
- Security Management: Create scanning protocols for AI-generated code vulnerabilities
- Software Inventory: Track AI tools used in development workflows and their outputs
- Developer Support: Provide training on responsible use of AI coding assistants
- Community Engagement: Engage with AI tool vendors on open source compatibility and ethics
#### Community
- Maintainers: Develop protocols for reviewing AI-generated patches and pull requests
- Contributors: Create educational resources on when AI coding tools are appropriate
- Ecosystem: Foster discussion about AI's role in sustainable open source development
#### WordPress
- Core: Evaluate AI tools that could assist with accessibility audits and performance optimization
- Documentation: Explore AI assistance for documentation generation while maintaining human oversight
- Community: Create forums for discussing responsible AI use in WordPress development
- Training: Develop curricula covering AI tools in WordPress development workflows
- Plugins: Establish review criteria for AI-generated plugin code submissions
## 🧪 Hypotheses
- Personal AI coding tools will follow the same adoption pattern as early web technologies: worker-driven rather than management-mandated
- The most successful AI applications will be highly specific, idiosyncratic tools that serve individual needs rather than mass markets
- Corporate AI initiatives will face resistance similar to early technology implementations that reduced user agency
- Local AI models will become more important as users prioritize control over their tools and data
- The current AI bubble will deflate, leaving behind useful personal automation tools as residual value
- Vibe coding will create a new category of "vernacular software" that is powerful for individuals but unsuitable for production
### 📊 Methodology
- Longitudinal study tracking adoption patterns of AI tools in personal vs. corporate contexts
- Qualitative analysis of user-generated AI applications and their sustainability over time
- Economic analysis comparing personal AI tool creation costs with traditional software development
- Comparative study of AI tool adoption versus historical technology adoption patterns
- Survey research on user preferences for local versus cloud-based AI tools
### ❓ Questions
- How can we distinguish between beneficial personal AI use and problematic production AI deployment?
- What technical and social mechanisms prevent the consolidation of vernacular AI tools into corporate platforms?
- How do we maintain the "deplatformed from inception" nature of personal AI tools?
- What would sustainable funding models for local AI development look like?
- How can open source communities manage AI-generated contributions without stifling innovation?
### 🔮 Future Research
- Long-term sustainability of local AI models versus cloud-based services
- Evolution of programming education in the age of AI coding assistants
- Impact of AI tools on creative and technical agency among non-programmers
- Development of quality assessment frameworks for AI-generated code
- Social dynamics of AI adoption in collaborative software development
### 💡 Implications
**Practice**: Developers need frameworks for appropriately integrating AI tools while maintaining code quality and security. Personal automation becomes more accessible to non-technical users.
**Policy**: Regulatory frameworks must distinguish between personal AI use and production AI deployment. Open source governance structures need updating for AI-generated contributions.
**Education**: Programming education must balance AI assistance with fundamental skill development. Digital literacy should include understanding AI capabilities and limitations.
**Security**: New vulnerability classes emerge from AI-generated code. Security practices must evolve to address both human and AI-generated threats.
**Economic**: The AI industry's focus on programmer replacement may miss more sustainable markets in personal productivity tools.
### 🧠 Learning Styles Assessment
Reading level estimate: Flesch-Kincaid Grade Level 12-13 (college level). Complex sentence structures and technical vocabulary require advanced reading skills.
#### 👀 Visual learners
**Present**: Historical timeline of computing development, contrasts between corporate and vernacular AI
**Improve**: Diagrams showing user agency spectrum, flowcharts distinguishing appropriate AI use cases, before/after screenshots of personal automation tools
#### 🔊 Aural learners
**Present**: Narrative storytelling structure, clear argumentation flow
**Improve**: Audio version of key concepts, podcast discussion of main themes, spoken examples of vibe coding applications
#### 🗣️ Verbal learners
**Present**: Rich analogies (cyberpunk, street finding uses), defined technical terms, historical context
**Improve**: Glossary of AI and programming terms, more detailed explanations of technical concepts like APIs
#### 🧩 Kinesthetic learners
**Present**: References to hands-on computing experiences, practical examples
**Improve**: Step-by-step tutorials for personal AI tools, interactive demonstrations of vibe coding, hands-on exercises comparing AI outputs
#### 🧠 Logical learners
**Present**: Clear argumentation structure, cause-and-effect relationships, economic analysis
**Improve**: Decision trees for appropriate AI use, comparison tables of different AI applications, logical frameworks for evaluation
#### 🧑🤝🧑 Social learners
**Present**: Community examples (Archive of Our Own, open source projects), collaborative development themes
**Improve**: Discussion prompts about AI ethics, case studies of community responses to AI tools, collaborative exercises
#### 🧘♀️ Solitary learners
**Present**: Personal computing history, individual project examples
**Improve**: Self-reflection questions about technology use, personal assessment tools for AI adoption, individual exploration exercises
#### 🌿 Naturalistic learners
**Present**: Organic technology adoption patterns, ecosystem metaphors for software development
**Improve**: Analogies to natural selection in technology adoption, environmental impact discussions of AI computing
### 📊 Accessibility Assessment
#### Screen reader semantics
**Present**: Clear heading structure, descriptive link text
**Improve**: Alt text for any embedded images, ARIA labels for complex navigation, table headers for data comparisons
#### Keyboard navigation
**Present**: Standard web navigation patterns
**Improve**: Skip links for long content sections, logical tab order for interactive elements
#### Contrast and readability
**Present**: High contrast black text on white background as noted in site description
**Improve**: Verify contrast ratios meet WCAG 2.2 AA standards, consider increased line height for long-form content
#### Text resizing and reflow
**Present**: Responsive design elements
**Improve**: Test at 200% zoom for readability, ensure no horizontal scrolling required
#### Media and captions
**Present**: Text-based content without embedded media requiring captions
**Improve**: Any future video content should include captions and transcripts
### 🌍 DEIB Assessment
#### Cultural breadth
**Present**: Global perspective on technology development, awareness of different user needs
**Improve**: More examples from non-Western technology traditions, recognition of diverse computing experiences
#### Inclusive language
**Present**: Avoids technical jargon without explanation, uses accessible metaphors
**Improve**: Plain language summaries of complex concepts, expanded glossary for technical terms
#### Neurodiversity support
**Present**: Logical argument structure, clear section headings
**Improve**: Content warnings for rapid topic changes, consistent formatting patterns, summary boxes for key points
#### Representation
**Present**: Diverse examples of technology users and creators
**Improve**: More balanced representation across different demographics in examples and case studies
#### Psychological safety
**Present**: Respectful critique of technology industry, acknowledges different perspectives
**Improve**: Explicit recognition of diverse viewpoints on AI adoption, content warnings for industry criticism
### ✅ Logical Validity
#### Strong reasoning
**Analogical reasoning**: The comparison between early web adoption (worker-driven) and current AI adoption (boss-mandated) effectively illustrates adoption pattern differences and predicts resistance.
**Causal reasoning**: The connection between iPad's consumption-focus and reduced user innovation demonstrates how technology design affects user agency.
**Inductive reasoning**: Using specific examples (plane-tracking app, Wreccer project) to support broader claims about vernacular AI applications strengthens the argument through concrete evidence.
#### Fallacies
**Appeal to tradition**: While drawing on cyberpunk philosophy and early internet history provides valuable context, the argument occasionally implies that newer approaches are inherently worse without fully examining their merits.
**False dichotomy**: The framing sometimes presents a binary choice between corporate control and user agency, when hybrid models might exist.
#### Tone of voice
**Conversational yet authoritative**: "I could see where she was coming from" demonstrates approachability while maintaining expertise. **Critical but constructive**: Technology criticism includes positive examples and solutions rather than pure negativity.
### 🛡️ DARVO Analysis
No DARVO patterns detected. The article critiques industry practices without reversing victim/offender dynamics, attacking critics, or denying legitimate concerns. The critique focuses on systemic issues rather than personal attacks.
## 🎯 Conclusion
Doctorow's analysis reveals a crucial distinction between AI as a tool for personal empowerment versus corporate control. The real value lies in enabling individuals to create idiosyncratic solutions for their specific needs, continuing the tradition of user-controlled technology. However, the industry's focus on replacing skilled workers and centralizing control threatens to squander this potential. For practitioners and organizations, the key is recognizing when AI serves genuine user needs versus when it extracts value while reducing agency. The future of beneficial AI development may depend on preserving space for vernacular, personal applications while resisting the pressure to scale everything into production systems or corporate platforms.
## 🔄 Refactor Appendix — AI and amateurism
### 📚 Literature Atoms (H3)
#### Personal AI tools enable vernacular software creation
> [!quote]
> The stuff we make on our own computers, for ourselves, is deplatformed from its inception. It's part of the life we can build in technology's "shadowy corners" that we used to just call "technology."
**Context & use**:
- Core argument about personal vs. corporate AI applications
- Distinguishes between individual empowerment and platform dependency
**Tags**: #xref #vernacular-technology
**Highlight**: Direct quote from article main content
<!-- atomicity: 1 idea ✓ -->
#### Corporate AI focuses on programmer replacement over user empowerment
> [!quote]
> To keep the bubble inflated, AI hucksters must promise massive economic returns to the technology. They want investors to believe that vibe code is about to replace working programmers.
**Context & use**:
- Economic analysis of AI industry motivations
- Explains mismatch between technology potential and market positioning
**Tags**: #ai-bubble #xref
**Highlight**: Direct quote from economic analysis section
<!-- atomicity: 1 idea ✓ -->
#### User-controlled tools enable better solutions than expert design
> [!quote]
> If "epistemic humility" means anything, it means acknowledging that no amount of "requirements gathering" can capture the needs of people totally unlike yourself as faithfully as those users can capture their own needs.
**Context & use**:
- Philosophical foundation for user-controlled technology
- Supports argument for personal AI tools over corporate solutions
**Tags**: #xref #user-agency
**Highlight**: Direct quote from main argument
<!-- atomicity: 1 idea ✓ -->
### 🧠 Permanent Candidates (H4)
#### Vernacular technology serves individual needs better than scaled solutions
**Claim**: Personal software tools that are "unsuited for production code" can be more valuable than professionally developed alternatives because they directly serve specific user needs without compromise.
**Evidence pointer**: [[#Personal AI tools enable vernacular software creation]]
**Scope & conditions**: Applies to individual productivity tools and personal automation; breaks down when reliability, security, or multi-user coordination becomes critical.
**Links**:
- [[User agency in software design]] — because extends the principle of user control
- [[Technology adoption patterns]] — because contrasts with corporate-driven adoption
<!-- atomicity: 1 idea ✓ -->
### ✏️ Fleeting Queue
- [F1] Investigate Wreccer project architecture and API design principles — https://github.com/wreccer #fleeting #inbox
- [F2] Research historical patterns of technology adoption: worker-driven vs. management-mandated — cyberpunk literature + early web history #fleeting #inbox
- [F3] Compare economics of personal automation tools vs. commercial AI services — cost analysis needed #fleeting #inbox
### → Things 3
- [ ] Research Wreccer project's approach to user-customizable interfaces — development patterns #things
- [ ] Document examples of successful personal automation tools for blog post — writing #things
<!-- deep-read:sections end -->
## Highlights
- The stuff we make on our own computers, for ourselves, is deplatformed from its inception. It's part of the life we can build in technology's "shadowy corners" that we used to just call "technology.
- Note: 📚 Personal AI tools enable vernacular software creation · - Core argument about personal vs. corporate AI applications - Distinguishes between individual empowerment and platform dependency · #atom/literature/h1
- To keep the bubble inflated, AI hucksters must promise massive economic returns to the technology. They want investors to believe that vibe code is about to replace working programmers.
- Note: 📚 Corporate AI focuses on programmer replacement over user empowerment · - Economic analysis of AI industry motivations - Explains mismatch between technology potential and market positioning · #atom/literature/h2
- If "epistemic humility" means anything, it means acknowledging that no amount of "requirements gathering" can capture the needs of people totally unlike yourself as faithfully as those users can capture their own needs.
- Note: 📚 User-controlled tools enable better solutions than expert design · - Philosophical foundation for user-controlled technology - Supports argument for personal AI tools over corporate solutions · #atom/literature/h3
- Vernacular technology serves individual needs better than scaled solutions
- Note: [[User agency in software design]] — because extends the principle of user control · - [[User agency in software design]] — because extends the principle of user control · #atom/zettel/z1
- Investigate Wreccer project architecture and API design principles — https://github.com/wreccer
- Note: ✏️ Fleeting capture #atom/fleeting/f1
- Research historical patterns of technology adoption: worker-driven vs. management-mandated — cyberpunk literature + early web history
- Note: ✏️ Fleeting capture #atom/fleeting/f2
- Compare economics of personal automation tools vs. commercial AI services — cost analysis needed
- Note: ✏️ Fleeting capture #atom/fleeting/f3