OpenAI pulled Study Mode from ChatGPT without telling anyone. Users on Hacker News spotted the feature had vanished, and there's been no formal announcement explaining why. The mode guided ChatGPT to use Socratic-style questioning for studying. Now it's just gone.
Study Mode was just a system prompt, a hidden instruction injected at the start of your conversation that told the underlying model how to behave. That's why OpenAI could remove it silently. No retraining, no infrastructure changes, just deleting a text configuration. These modes work through pure prompt engineering, persistent text directives that shape responses without touching model weights.
For developers building specialized AI apps, this is a familiar anxiety cycle. One commenter who built listendock.com, a study-focused tool, initially worried when OpenAI launched features overlapping with their product. Features that appear and vanish without notice make poor foundations. If OpenAI can't commit to keeping a study mode alive, there's space for focused apps that actually stick around and serve users consistently. The codebases that work best for AI agents probably won't look like what we'd write for humans. Flat hierarchies beat abstractions, and your CLAUDE.md file matters more than your linter.