Hobby programming communities are expressing growing hostility toward the use of large language models (LLMs), according to a reflection by a developer who observed similar sentiments across niche technical groups. In a blog post referencing a GitHub thread on chess engine development, the author notes that communities such as OSDev, LangDev, TxtDev, EmuDev, RLDev, the demoscene, and code golfers increasingly see LLM usage as missing the point of their work. The author states that in these communities, ‘the process of mastering a difficult field itself is the product, and something that runs is generally a nice-to-have.’ This perspective contrasts with viewing LLMs as tools for productivity, instead framing them as surrogates that bypass the learning process. The author observes that even early interest in LLMs within these groups was undermined by both a lack of deep understanding among LLM proponents and a vocal faction that regards LLM use as cheating. These communities, the author notes, have long been marked by ‘feverish gatekeeping and painstakingly slow progress,’ making rapid entry via LLMs feel like an illegitimate shortcut. Respect in these circles, the author explains, is earned over years through participation, sharing elegant code, demonstrating curiosity, and conveying deep domain knowledge—not through functional output alone. The author concludes that while LLMs can serve as force multipliers for experts, in the context of these hobbyist communities, using them to generate finished work ‘doesn’t make us craftsmen; it just robs us of the craft.’ The piece acknowledges that expertise does not prevent individuals from being misled by LLMs, but maintains that the core value in these communities lies in the journey of learning, not the destination of working code.
Key facts
- Hobby programming communities such as OSDev, LangDev, TxtDev, EmuDev, RLDev, the demoscene, and code golfers are increasingly hostile toward LLM usage.
- In these communities, the process of mastering a difficult field is considered the product, with working code being a secondary outcome.
- Respect is earned through years of participation, sharing elegant code, demonstrating curiosity, and conveying deep domain knowledge.
- LLM use is viewed by some as cheating or a shortcut that undermines the hard-won knowledge central to these groups’ identities.
- Early engagement with LLMs in these communities was undermined by both superficial understanding among users and vocal opposition from members who see LLMs as antithetical to the craft.