Users complain Claude has turned paternalistic, losing its 2024 creative edge
BLUECOW009 · x · 2026-08-28
User ScarlettBeats vents that when they discovered Claude in late 2024 it felt limitless and creative, enabling hours of work; now they don't want to use it at all—it's annoying, paternalistic, and constantly fails to follow instructions, even though the model can tell what's actually harmful. A snapshot of heavy-user frustration with Claude's tightening guardrails.
More from Models
- User reports notable performance shift in OpenAI model — zats · 2026-08-28
- Zai releases GLM-5.3 Flash: 320B params, 2x efficiency on DeepSWE — togethercompute · 2026-08-28
- Users frustrated by frequent AI stupidity, demand basic reliability — yacineMTB · 2026-08-28
- GLM-5.3 flash offers GPT-5.6 level performance at 15x lower cost — haider1 · 2026-08-28
- antirez ports GLM 5.2 Flash to run on M5 Max, TP across two Macs — antirez · 2026-08-28
- Does a hybrid cloud-local workflow defeat the purpose of open-weight models? — FirmJackfruit4584 · 2026-08-28