Tell HN: OpenAI silently removed Study Mode from ChatGPT

OpenAI quietly axes Study Mode — fans wail, skeptics shrug, YouTube hype gets roasted

TLDR: OpenAI quietly removed ChatGPT’s Study Mode, and the crowd split fast: some say it was just a preset you can recreate, others gripe that useful tools and quality are being trimmed for product metrics. Jokes flew about YouTube hype and Google-style feature culls, with security and simplicity cited as possible reasons.

OpenAI quietly pulled ChatGPT’s “Study Mode,” and the internet did what it does best: argue about it with flair. One camp is eye-rolling, led by users like brumar insisting the feature was just a fancy label — “basically a preset prompt.” Another camp is mourning their AI tutor, with autodidacts fuming that a useful mode got chopped for product metrics. Some suspect a “growth-first” cleanup: fewer buttons, fewer headaches. Others whisper about a security angle — exposed prompts can be a jailbreak surface — but it’s speculation.

The hottest spice? People saying ChatGPT itself feels bland and salesy lately — “PR voice,” “confidently wrong,” and “middle-management vibes.” Meanwhile, el_io shrugs: can’t you just ask it to behave like a study coach anyway? Entrepreneur janpmz sees opportunity: niche apps might win where big platforms keep pruning. And altmanaltman dunks on those viral thumbnails — “OPENAI CHANGED STUDYING COMPLETELY…” — with the punchline: “Guess studying changed it.”

The thread’s mood swings from grief to gallows humor. CatDeveloper_ likens the move to Google-style spring cleaning, while others share DIY hacks to recreate the mode in one line. The verdict? No consensus — just peak Hacker News energy, a swirl of memes, metrics theories, and salty nostalgia. Dive into the back-and-forth on Hacker News.

Key Points

  • The article states that OpenAI removed the “Study Mode” feature from ChatGPT.
  • The author proposes product strategy as a likely reason, citing feature surface area and retention metrics.
  • The piece raises a security consideration: exposing or inferring system prompts could expand jailbreak and prompt-injection risk.
  • The author reports a perceived decline in ChatGPT’s output quality, describing responses as boilerplate and overly confident.
  • The author claims other large language models do not show the same behavior, suggesting a potential shift in ChatGPT’s tone or target audience.

Hottest takes

"this \"mode\" was just a system prompt" — brumar
"I guess studying changed it" — altmanaltman
"kinda like how google owuld remove some features at random.." — CatDeveloper_
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.