Desperately Seeking Software ![rw-book-cover](https://aboard.com/wp-content/uploads/2025/08/gpt5-wide.png) ## Metadata - Author: [[Paul Ford]] - Full Title: Desperately Seeking Software - Category: #articles - Document Note: --- tags: - #AI - #Software - #TechIndustry Author: Paul Ford Source: Aboard Newsletter Reference: [Desperately Seeking Software: A review of ChatGPT 5](https://aboard.com) (Aug 13, 2025) date created: 2025-08-22 21:55 --- # Title: Desperately Seeking Software – A Review of ChatGPT 5 ## 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-diversity-equity-inclusion-belonging-assessment) - [✅ Logical Validity](#-logical-validity) - [🔍 DARVO Analysis](#-darvo-analysis) - [🔍 SEO Evaluation](#-seo-evaluation) - [🎯 Conclusion](#-conclusion) --- ## 📝 Summary: Paul Ford reviews **ChatGPT 5**, framing it as an ambitious but chaotic release. While OpenAI has consolidated its product family into a unified interface with “routing,” the execution often feels unfinished. Ford notes that Codex (for developers) works well, but other features like “agents” and video generation are half-baked. The overarching theme: OpenAI needs to transition from experimentation to proper product development if it hopes to sustain trust and usability at scale. --- ## 📖 Main Content: Ford evaluates ChatGPT 5 across six dimensions: What’s this? Who’s it for? What’s new? Does it matter? What’s missing? What’s it cost? - **What’s This?** A unification of OpenAI’s sprawling product line into a single conversational interface via routing. - **Who’s it For?** Everyone, theoretically. But the lack of clarity in vertical vs horizontal apps undermines focus. - **What’s New?** Routing simplifies choice of model, collapsing multiple LLMs into ChatGPT 5 and “Thinking” mode. However, many features (e.g., agents, Sora integration) break or confuse users. - **Does it Matter?** Yes, consolidation matters, but execution gaps resemble software bloat eras (Windows Vista, MacOS 9). - **What’s Missing?** Onboarding, clarity, documentation, vertical-specific apps, and transparency about limitations. - **What’s it Cost?** Free tier + $20/mo Plus; $200/mo Pro; enterprise plans and cheap API tiers. Ford closes by calling for OpenAI to embrace its role as a **software company**, not just an experimental lab, stressing that AI must integrate into real-world workflows. --- ### 🗒️ Notes - OpenAI’s naming chaos (4, 4o, 4.5, DALL·E, Sora) → ChatGPT 5 aims to consolidate. - Routing = simplified UX but launched buggy (miscounting letters). - Desktop app: icons for uploading docs, research, agents, thinking mode. - Examples of failure: - Hippo video glitching. - “Agents” making promises they can’t keep. - Sciatica exercise cats → cute but useless. - Examples of success: - Codex onboarding (GitHub repo → PR). - Synth patch planning = improved reasoning + diagrams (though still flawed). - Critique: Product lacks **self-awareness, onboarding, and vertical clarity.** - Ford invokes history: Bill Gates fuming at broken Movie Maker installs = the kind of grit OpenAI needs. --- ## 📖 References: - Paul Ford, *Aboard Newsletter* (Aug 13, 2025). - OpenAI livestream, ChatGPT 5 launch. - Bill Gates “Movie Maker” memo (2003) as analogy. --- ## 🗣️ Paraphrase: 1. Original: “This release is important, because it shows OpenAI trying to make a single, consolidated platform—but if you want to be a useful mega-platform, could you button it up a little?” → Rephrased: OpenAI’s consolidation effort is meaningful, but the company must mature its product into something polished and reliable if it wants platform credibility. 2. Original: “The fool doth think he is wise, but the wise man knows himself to be a fool… and that’s a product problem.” → Rephrased: ChatGPT’s biggest weakness is its lack of self-awareness—it can’t explain its own features or guide users, making it more confusing than empowering. --- ## 📰 Newsletter: 🔥 **This Week in WordPress & Beyond: Software Lessons from ChatGPT 5** Paul Ford’s review of **ChatGPT 5** hits home for those of us building WordPress and open-source tools. His verdict? The software is powerful but messy—lacking onboarding, transparency, and proper vertical applications. Why should WordPress folks care? Because these same lessons apply to **block editor adoption, plugin development, and contributor engagement**. Without polish and clarity, even groundbreaking tools risk alienating users. Key takeaways: - Simplify naming and interfaces. - Provide proper onboarding & documentation. - Respect vertical contexts (dev tools, health, publishing). 👉 Full review: [Aboard Newsletter](https://aboard.com) --- ## ✅ Action Items: - Relate ChatGPT’s **onboarding failures** to WordPress contributor experience; push for **clearer entry points**. - Advocate for **vertical clarity**: docs, patterns, dev tools must be context-specific. - Position this critique in OSPO/GoDaddy strategy: AI in open source requires **trust, reliability, and transparency**. - Use Ford’s framing (Vista-era analogy) as a cautionary tale for **sustainable WordPress innovation**. --- ## 🧪 Hypotheses: - H1: Clear onboarding increases user retention across AI and OSS tools. - H2: Vertical-specific AI apps (e.g., Codex for devs) outperform generalist models in adoption. - H3: Lack of documentation directly correlates with user frustration and churn. ### 📊 Methodology: - Compare onboarding flows in ChatGPT vs WordPress. - Conduct surveys on trust levels tied to product transparency. - Track adoption of vertical plugins vs general-purpose tools. ### ❓ Questions: - Should OSS tools emulate Codex-style onboarding? - How much tolerance do users have for “half-launched” features? - Can transparency about limitations boost trust, even if functionality lags? ### 🔮 Future Research: - Explore **cross-comparison of AI onboarding vs OSS onboarding**. - Study the impact of **naming clarity** on adoption. - Investigate **trust gaps in AI vs traditional software**. ### 💡 Implications: - AI/OSS need to move beyond hype → toward sustainable, **trustworthy products**. - WordPress governance can learn from OpenAI’s missteps: clear, consistent user journeys. --- ## 🧠 Learning Styles Assessment: - Visual: ✅ Screenshots, icons, diagrams (though flawed). Needs better infographics. - Aural: ❌ No audio/podcast integration. Could use audio reviews. - Verbal: ✅ Narrative review format with strong analogies. - Kinesthetic: ❌ No hands-on tasks provided. - Logical: ✅ Structured Q&A format. - Social: ✅ Newsletter fosters dialogue. - Solitary: ✅ Reflection-heavy critique. - Naturalistic: ❌ No environmental framing. Readability: College-level (12th–14th grade). --- ## 📊 Accessibility Assessment: - Images provided but lacking alt-text clarity. - No audio versions. - Text is navigable but long; could benefit from shorter chunks. - Needs stronger contrast in embedded screenshots. --- ## 🌍 DEIB (Diversity, Equity, Inclusion, Belonging) Assessment: - Cultural scope: Narrow; assumes tech-savvy Western readers. - Inclusive language: Neutral, but jargon-heavy. - Neurodiverse support: Needs shorter, more scannable structure. - Representation: Focused on AI insiders, not broad user groups. - Safety: Tone = critical but not hostile. --- ## ✅ Logical Validity: - Strengths: - **Causal reasoning** (lack of onboarding → user failure). - **Consistent analogies** (Vista/Movie Maker as precedent). - **Balanced critique**: acknowledges successes like Codex. - Weaknesses: - Occasional **hasty generalizations** (“no one reviews with words anymore”). - Slight **appeal to emotion** (frustration with hippo glitches). Tone: Witty, skeptical, - Summary: If you want to be a true platform for the future, you have to actually make the thing work. - URL: https://aboard.com/desperately-seeking-software/ [the aboard newsletter](https://aboard.com/newsletter/) A review of OpenAI’s latest big release, ChatGPT 5. ![Image of "AI Stars" and a half symbol against a soft pink and blue background.](https://aboard.com/wp-content/uploads/2025/08/gpt5-tall.png)Chat, how would you rate yourself?, Desperately Seeking Software, Paul Ford, 01k3215ycc1gyrjkra4vtgvxdy.