Compute likely tied up in pretraining, but Claude dev experience keeps getting worse
ATTlKA · x · 2026-09-06
Following up on his Claude Max quota complaint, the author says he has no doubt the compute is tied up in pretraining "something presumably magnificent," but the current developer experience is poor — even the $200 plan can no longer sustain normal usage.
Related event: Users Complain Claude Max 20x Quota Quietly Shrinking(2 posts)→
More from Models
- Diffusion LLM bias rankings shift with sampling configs, persist after debiasing — Bollegala · 2026-10-23
- Running ~100 DeepSeek v4.1 flash agents in parallel for under $20/month — zack_overflow · 2026-10-23
- Researcher laments open-source audio models are surprisingly subpar, even representation ones — gowthami_s · 2026-10-23
- User claims they generated video with "Claude Opus 5.5", unannounced model name surfaces — jwt0625 · 2026-10-23
- Security researchers push back as OpenAI restricts Astra access; BlackHat 2027 mocked as 'cyber model' bazaar — cyb3rops · 2026-10-23
- Mistral Large 4 Leaked on Hugging Face: 1T Params, 49B Active, Open Weights Oct 31 — mervenoyann · 2026-10-23