PhD researcher says ChatGPT's constant redesigns are eroding his trust in the product
jaydeelive01 · reddit · 2026-10-04
A PhD researcher who relies on ChatGPT for writing, coding, reading papers and project management posted a detailed critique: OpenAI ships new paradigms faster than users can rebuild their mental models.
- Redesign fatigue: After Projects came Work, local/cloud, and now Spaces and Pages — he jokes the Work vs Codex distinction is a Navier-Stokes problem, and says Pages is basically the old Codex UI with persistent storage, worse than the Chat + Codex side-by-side setup from three years ago.
- Trust problem: Every launch makes him wonder if regular Chat is being deprecated; web, desktop, local and cloud all behave differently across Windows/Mac/Linux/iOS, and he sometimes asks a Pro model what ChatGPT can actually do this month.
- Half-baked features: a collaborative writing tool without tracked changes.
- ADHD perspective: AI helps him follow thoughts instantly, but when the tool itself demands constant relearning, he ends up exploring ChatGPT instead of working.
His verdict: he won't move his workflow into Spaces, keeping notes in Obsidian, papers in a reference manager, manuscripts in Word and code in repos — with ChatGPT working across them — until OpenAI keeps working features stable and warns users before major changes.
More from Models
- Mystery Model 'Gestalt' Tops New Frontier Design Benchmark by 214 Points — soleio · 2026-11-19
- mrfakename releases abliterated GLM-5.3-Flash and Qwen3.8-Flash-Next — mrfakename0 · 2026-11-19
- Physician: AI competition now measured in weeks, GPT-6 versions 7 days apart — ramos_casals · 2026-11-19
- micro1's flow qc model screens training data for benchmark leaks before release — Exp_Mark · 2026-11-19
- Leaker predicts new DeepSeek paper this week, scheduled for November 11 — teortaxesTex · 2026-11-19
- Gemini 4 Neon rollout begins: leaked UI shows 3-tier Gemini 4 lineup with Carbon locked — testingcatalog · 2026-11-19