Is it all just vapourware?
kirahowe.com
August 9, 2026
3 min read
🔥🔥🔥🔥🔥
48/100
Summary
Concerns are raised about the prevalence of vapourware in the AI and agent software space, highlighting significant friction within the ecosystem. The lack of systematic review, quality assurance, and usability testing is noted as a major issue impacting the development of these technologies.
Key Takeaways
- Many AI and agent software products are perceived as vapourware due to poor user experience and lack of systematic quality assurance processes.
- The cost of writing code has decreased significantly, leading to an oversaturation of startups with unrefined execution and inadequate consideration for user experience.
- The author experienced significant friction while using the ONA platform, including login issues and inefficient task management, which detracted from the intended benefits of the tool.
- Marketing claims of agentic development are seen as exaggerated, as the companies promoting these tools do not demonstrate their effectiveness in their own software.
Community Sentiment
Positives
- Some engineers report that LLMs can write code much faster and handle edge cases better than they could on their own, revealing a valuable partnership in development.
- Prominent contributors like antirez and Mitchel Hashimoto showcase that skilled engineers can leverage LLMs effectively, hinting at potential productivity gains when used correctly.
Concerns
- The overwhelming majority of LLM-generated contributions to open source projects are criticized as low quality, raising doubts about their actual productivity benefits.
- Many commenters express skepticism about the real-world effectiveness of LLMs, citing numerous instances of incorrect outputs and poor code quality, leading to frustration.
