Why Gumroad Didn’t Choose htmx: A Critical Reflection on AI Tool Dependency and the Balance Between Innovation and Automation
Recently, the team at Gumroad shared an interesting article explaining why they chose not to use htmx for a new project called Helper, opting instead for Next.js. Among the main reasons was the fact that the React/Next.js ecosystem is larger and more mature, offering pre-packaged solutions to common problems, whereas with htmx they often had to reinvent solutions (making it more suitable for less complex projects). Additionally, htmx showed scalability limits as the project grew, turning initial simplicity into an obstacle in managing complex states. Lastly, due to greater developer familiarity, Next.js was perceived as more intuitive and natural compared to htmx.
But, one of the main reasons that influenced their decision, and one that particularly struck me, was the greater familiarity of AI tools with Next.js compared to htmx, due to the wider availability of open-source training data. This made development with Next.js faster and problem-solving easier, thanks to the vast amount of accessible resources and knowledge.
AI and Tooling Support: It’s worth noting that AI tools are intimately familiar with Next.js and not so much with htmx, due to the lack of open-source training data. This is similar to the issue Rails faces. While not a dealbreaker, it did impact our development speed and the ease of finding solutions to problems. When we encountered issues, the wealth of resources available for React/Next.js made troubleshooting much faster.
One of the most intriguing points of their reasoning is, therefore, the influence of AI tools in choosing the technology to use, which raises a fundamental question: are we allowing AI to guide our technological choices, thereby limiting the potential for innovation?
The Risk of Technological Stagnation
Behind this choice, perfectly legitimate in a production context where speed and reliability are crucial, lies a deeper issue that deserves critical reflection: is it right to choose a development language based on better support from AI tools? If this mindset takes root, it could lead us toward a future where the creation of new programming languages or experimental technologies becomes an increasingly rare event. Why? Because everyone will be afraid to embrace new technologies if they haven’t yet accumulated enough data to be leveraged by modern AI tools, or if they don’t offer sufficient resources for quickly solving production problems.
The risk is that the technological ecosystem could become stagnant, dominated solely by the most popular technologies, which enjoy a vast community and support, but which may not always be the most efficient or elegant solution. The introduction of new paradigms, such as those promoted by new programming languages like htmx, is essential to stimulate innovation and overcome the limitations that existing technologies may present. Without this ability to evolve and introduce new ideas, the entire sector could slow down, risking being stuck in solutions that, while widely supported, no longer adequately respond to the increasingly complex and varied demands of the technological world.
Standardization and Its Effects
In Gumroad’s case, the choice of Next.js was driven by efficiency and the ability to fully leverage AI support to speed up development. But what happens if everyone follows this path without considering alternatives? If companies and developers consistently choose technologies based on AI familiarity, we could witness extreme standardization, where new ideas struggle to take root. This phenomenon of “technological laziness” could limit not only the ability to innovate but also the possibility of diversifying technological solutions, thus stifling both creativity and innovation, key elements for tackling future challenges. Emerging technologies may never have the chance to prove their worth, creating an ecosystem dominated by a few established players.
The Importance of Exploring Uncharted Paths
This situation makes us reflect on a vicious cycle that could emerge: if only already established technologies are adopted, they will become increasingly stronger, while emerging technologies will fail to find a gap to prove their worth. Dependency on AI tools — and the data on which they are trained — could ultimately penalize innovation and technological diversity, which have historically been the main drivers of progress in the software development world. If we continue to use only what is supported by AI because it is more convenient and faster, we risk losing the opportunity to discover alternative approaches that might be better suited to certain contexts and specific problems.
In any field, whether scientific or technological, true innovation comes from exploring uncharted paths and having the courage to experiment, even when there is no complete manual or AI tool ready to provide all the answers. AI tools have become inseparable companions for developers, offering code suggestions, problem-solving, and optimizations. However, this convenience comes at a cost. Relying too much on AI can reduce developers’ ability to think critically and creatively. This could lead to a superficial understanding of technologies and a lack of fundamental skills. Depth of knowledge and the ability to solve complex problems without assistance are essential for innovation. If developers become mere executors guided by AI, technological progress could slow down significantly.
Redefining AI’s Role in Development
It is therefore crucial to redefine the role of artificial intelligence in the creative and decision-making process. Developers should be encouraged to explore new technologies regardless of AI support, using AI as an aid, not a guide. AI should be a tool in the service of innovation, not a limitation.
It is therefore appropriate to balance the desire for immediate efficiency with the need to invest in the future of the industry. If, as a community, we stop exploring new technologies just because they are harder to use or less supported, we risk losing the opportunity to create better tools that are more suited to future needs. The choice to embrace new technologies may seem riskier in the short term, but as we know, true progress does not happen when everyone follows the easiest path, but when someone chooses to explore the unknown, accepting the possibility of failure but also of finding something extraordinary, ensuring the resilience of our ecosystem.
Companies and communities should promote the adoption of emerging technologies by investing in training and creating environments that value experimentation. Encouraging knowledge sharing, organizing hackathons on new technologies, and integrating emerging languages and tools into academic curricula can make a difference.
Originally published at Levysoft.
