May 4, 2026
Talk LLMs to me… or please don’t
Let's Talk about LLMs
Even the comments are tired of the LLM debate—but the fights are getting louder
TLDR: The article argues that chatbot-style software may be important, but probably isn’t a magic fix for all programming problems. The comments stole the show: some readers are totally exhausted by endless LLM debates, while others insist this changes everything and old rules no longer apply.
A blog post called "Let's Talk about LLMs" tried to do the calm, thoughtful thing: define terms, avoid the fuzzy catch-all label of "artificial intelligence," and ask the big question hanging over the whole industry—are these chatbot-style systems a world-changing leap, a bubble, or just another overhyped tech moment? The writer leans on old software wisdom from Fred Brooks’ famous "No Silver Bullet" idea: there may be no single magic fix for making software work dramatically easier.
But the real fireworks were in the comments, where a chunk of the crowd basically yelled, "Please, not this again." One reply was hilariously blunt: "Actually can we not thanks." Another practically rolled its eyes through the screen, saying yet another personal essay on language models in 2025 is simply not interesting anymore. That boredom itself became the drama: not everyone is arguing about whether the tools are amazing—some are arguing that the discourse has become the most exhausting product of all.
Then came the counterattack. The pro-LLM camp insisted skeptics are missing the point completely: these tools don’t just write code faster, they can help think, research, and test faster too. One commenter swerved into full movie-trailer mode—"And then AI happened"—to declare that the old “no 10x productivity boost” rule may finally be dead. So the vibe is split between eye-roll fatigue, messianic hype, and a side dish of nerdy philosophy links. In other words: the internet is absolutely still talking about LLMs, even when half the room is begging it to stop.
Key Points
- •The article frames current debate over LLMs as unresolved, ranging from transformative productivity gains to hype or bubble scenarios.
- •It uses “LLM” instead of “AI” to keep the discussion focused on large language models rather than the broader and less precise AI label.
- •The article defines “LLM coding” broadly as any use of an LLM to generate program code, whether supervised by humans or not.
- •The discussion is explicitly limited to technology and programming, avoiding broader societal commentary.
- •Fred Brooks’ *No Silver Bullet* is introduced to argue that software development contains both essential and accidental difficulties, with memory management cited as an example of the latter.