Has Claude completely detached himself from language and creativity?
retained text
Hello ! I’m raising this as a question **not a criticism** based on something I’ve noticed recently while using Claude, a tool I’ve relied on for nearly two years now. I don’t want to rehash the points already discussed ad nauseam on Reddit; I’ve read plenty of opinions on the subject. I simply need to determine whether my subscription is still worthwhile for my specific use case moving forward. To elaborate: I’ve been using a subscription specifically the "x5" tier primarily for creative writing, character research and analysis, and dialogue drafting. In this process, I’ve invested time in creating .md (Markdown) files and developing detailed, precise instructions and "personas" tailored to each specific project. Along the way, I’ve frequently reevaluated my own writing, conducted further research into AI, explored scripting techniques, and studied how to craft effective prompts for platforms like ChatGPT and Claude essentially learning how to better communicate with an AI. In short, I conducted two years of research on this, involving the frequent use of models such as Opus 4.5 → 4.6, followed by 5.5, and currently Fable 5.2. It is clear that the way AI relates to the human element in terms of writing, creativity, emotional resonance, the diversity of dialogue, and the depth of reflection has changed radically. Even when I try to explain my vision, demonstrate what I’m looking for, or outline the intended creative path, nothing changes; the results consistently end up being mediocre not necessarily in terms of the model’s inherent capabilities, but when compared to how it performed two years ago. I don’t want to lay all the blame on the company there have been weeks where the results were generally better, even though I hadn’t changed a thing on my end. My main question now is: given the current state of things and the future direction of Anthropic, is Claude still well-suited for my specific needs creative work, language, and fiction writing? Does it still make sense to maintain a subscription? (For context, I’ve also tested ChatGPT, but the results are no longer satisfactory at all even though, back in the day, GPT-4o was perfect.) I’ve come to realize that the time I invest in this is increasingly becoming a source of annoyance and frustration feelings I don’t want to associate with a hobby or a side project. I don’t think it’s a matter of my own working method anymore; rather, it’s about the direction AI is taking in general specifically in fields like coding or other technical areas. It’s clear that some people on this subreddit seem very satisfied with how they use it, whereas I find myself struggling and frequently getting annoyed by the results, especially over the last few weeks. Anyway, that’s the summary. Now for my questions: 1. Do you think the creative and human aspect has been definitively abandoned by AI companies, including the team behind Claude? 2. Should I cancel my subscription, given how I use the service and what Anthropic currently offers its users? Thanks youu
document fetched 2026-11-20 00:22 · canonical https://reddit.com/r/ClaudeAI/comments/1y0brmt/has_claude_completely_detached_himself_from