Ditch Vibecoding: Buy AI-Enhanced Pro Software
After five months of AI experimentation, Matthew Yglesias rejects solo 'vibecoding' and wants established software companies to use AI coding tools for more, better, cheaper products sold to consumers.
Core Preference: Pro Companies Over Solo Vibecoding
Matthew Yglesias, after five months engaging with AI coding tools, explicitly rejects 'vibecoding'—casual, individual AI-assisted coding based on intuition rather than structure. Instead, he advocates for professionally managed software companies to integrate AI assistance into their workflows. This shift promises more software products that are simultaneously better in quality, higher in volume, and cheaper in price, directly benefiting end-users like him who purchase them.
The opinion counters the hype around personal AI coding empowerment, prioritizing scalable production by expert teams over DIY experiments. Simon Willison collects this quote on his blog, tagging it under AI, AI-assisted programming, vibe-coding, and agentic-engineering—highlighting the tension between unstructured 'vibe' approaches and more rigorous, agent-like engineering.
Implications for Builders
For developers and indie builders, this underscores a trade-off: while AI lowers barriers for solo creators, true impact comes from leveraging it at organizational scale. Companies adopting AI coding can outpace individuals, delivering reliable products faster. Readers experimenting with AI tools should evaluate if their 'vibecoding' yields shippable outcomes or if partnering with or building pro teams amplifies results more effectively.
This thin post serves primarily as a quotable contrarian take amid AI productivity discussions, without additional analysis from Willison.