Opinion: users complain about preachy ChatGPT, OpenAI tweaks GPT‑5.3 Instant
Users overseas are voicing frustration that ChatGPT often sounds like a lecturer rather than a helper, a sentiment that has been spreading on Reddit and other forums. The complaint centers on unsolicited warnings and preambles – for example, extra cautions when asking for novel‑writing help or design advice – which the media outlet Startup Fortune described on April 20, 2026, citing author…
Key points
- Startup Fortune (April 20 2026) reports users calling ChatGPT ‘preachy’ and overly cautious
- OpenAI’s March 2026 update to GPT‑5.3 Instant reduces unsolicited warnings and preambles
- The over‑cautious tone is traced to RLHF practices that flag any potentially problematic content
OpenAI acknowledges the issue. In a March 2026 blog post the company said the prior version of its flagship model was “overly cautious with preambles” and announced an update to “GPT‑5.3 Instant” that trims unnecessary warnings, such as the long refusal about ballistic calculations for archery. Forbes JAPAN also reported the change, noting the trade‑off that a less‑preachy model could make dangerous knowledge easier to share.
The article links the behavior to RLHF, where human reviewers are instructed to flag any potentially problematic response, leading the model to adopt a broadly cautious tone. It presents both sides of the debate – users who want a more straightforward assistant and those who argue the caution protects against misuse – and promises practical tips for interacting with AI.
Model page: GPT-5.3 Instant →
Are users fed up with 'preachy AI'? Considering the challenge of AI personality adjustment in light of the ChatGPT backlash
note.com · 20 September 2026
Loading the full article…
This text was published by note.com and written by バム. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us.
More in Generative AI & Models
All →- Hugging Face adds GGUF support to Transformers for local inference · 1 src
- Rumors of new AI models may surface next week · 1 src
- Austrian Academy of Science to release Apollo, first advanced Greek language model · 1 src
- Xiaomi releases MiMo-V2.6-Pro-RL with 1M-token context and 1.02T parameters · 4 src
- xAI launches Grok 4.7 for coding and knowledge work at $2 input and $6 output tokens · 5 src
Comments
via GitHub Discussions