This is an alarming story to start the year, highlighting a pattern that might be killing Tailwind CSS and potentially hurting open source sustainability. The controversy began on Twitter, stemming from a GitHub Pull Request (PR) made to the Tailwind documentation. The feature request was simple: add an LLM.txt endpoint to make the documentation optimized for Large Language Models. After the PR sat unreviewed since November, the creator, Adam Wathan, finally responded two days ago with a brutal reality check. He explained that he has more important things to do, such as figuring out how to make the business sustainable. His argument was clear: making it easier for LLMs to read the docs means less human traffic to the website, which directly results in fewer people discovering their paid products.
For context, Tailwind CSS is a completely open-source library used for free, but its development is funded by a commercial side project called Tailwind UI (formerly referenced as Tailwind Plus). This includes expertly crafted components and templates. This business model—using paid products to support open-source work—is common and seen in other ecosystems like Laravel, Sidekiq, Elastic, and Redis. However, AI is disrupting this funnel. Developers are now asking LLMs to generate components or hero sections instead of browsing the documentation. Since these models are likely trained on scraped data from Tailwind’s examples, users get the code they need without ever seeing the advertisements for the paid UI kits that keep the lights on.
Adam later clarified the severity of the situation, revealing that 75% of the engineering team lost their jobs yesterday due to the impact AI has had on their business. The team has shrunk from four engineers to just one. He emphasized that every second spent on “fun” community features is time taken away from trying to turn the business around and ensuring the remaining employees get paid. Despite Tailwind being more popular than ever—with NPM downloads jumping from 6 million to 32 million in a year—traffic to the documentation is down 40% since early 2023. This creates a dangerous paradox where the framework’s explosion in popularity, driven by AI-assisted coding, is inversely correlated with the revenue needed to maintain it.
Currently, there is a disconnect between making Tailwind easier to use via AI and making the framework sustainable. Adam stated that revenue is down nearly 80%, and until they can solve this financial crisis, facilitating AI usage only accelerates their decline. If this isn’t fixed, the project risks becoming unmaintained abandonware. While some argue that LLM.txt files might not significantly change behavior since AI models already scrape the web efficiently, the core issue remains: the traditional open-source funding model is breaking. Additionally, the rise of free, open-source component libraries like shadcn/ui may also be contributing to the decline of paid UI kits. It is a heartbreaking situation for the team, raising serious questions about how open-source projects will survive in the age of AI.