SI Agent News
RumorOpenAI

A ChatGPT user asks OpenAI for ready-made personality presets

A forum post proposes official, editable response-style templates for ChatGPT. It is a request, not a release, but it points at a real gap for beginners.

The short version

Why it matters

Many people never change how ChatGPT talks to them because they do not know how to describe the tone they want. Presets would lower that barrier. Until OpenAI says anything, the practical route is still writing a short set of custom instructions yourself.

A request, not a release

A post on community.openai.com, OpenAI's user forum, is titled "Official Personality Templates for ChatGPT." It is written as a feature request. The author proposes that OpenAI ship a set of official personality and response-style templates that people can pick and then adjust.

That distinction matters. Nothing in the post says OpenAI has built, tested or promised this. It is one user describing what they would like to see, and our own briefing earlier today flagged it as unconfirmed beyond the title. Read it as a signal of what users want, not as a product update.

Still, the idea is concrete enough to be worth a closer look, because it targets a common friction in how people use chatbots.

The problem it tries to fix

According to the post, anyone who wants ChatGPT to behave differently currently has to write their own custom instructions. Those are standing notes that tell the assistant how to respond, such as how formal to be or how much detail to give.

The author points out two costs. Writing them takes time, and it also takes some skill in putting a desired behavior into words. Plenty of people know they dislike a long-winded answer without being able to spell out what they would prefer instead.

The proposal is to skip the blank page. Offer ready-made options, and let people modify them rather than build from nothing.

Six styles on the wish list

The post names six starter templates:

Each one is a way of answering, not a character. A "verification-focused" style, for example, would presumably lean toward checking claims and questioning assumptions, while an "educational" one would lean toward explaining. The post does not spell out how any of them would behave, so those readings are our interpretation of the labels.

The author also argues that browsing such a menu would help people discover communication styles they could not have described on their own.

Editing, sharing and no sign-in

The request goes beyond a simple preset list. It asks for three further abilities.

Customization. A user could choose a template and change individual traits without writing instructions from scratch.

Sharing. People could pass their tuned templates to others. The author says this would encourage sharing and improving useful configurations.

Anonymous access. People who are not signed in could pick a template and use it during their session. The post presents this as a way to make ChatGPT more approachable for people without an account.

Each of these would be a bigger product decision than the templates themselves. Sharing raises questions about how templates are published and reviewed, and anonymous use would depend on how OpenAI treats logged-out sessions. The post does not address either.

Starting points, not restrictions

The most interesting part of the post is its design principle. Templates, the author writes, should be starting points. Users should be free to customize, combine or replace individual settings.

The post also says that when preferences conflict, for instance a request for both concise and educational answers, the user's own tweaks should settle it. The system should not block the combination.

That is a sensible stance, since real preferences rarely fit one label. Someone may want blunt answers for work emails and gentle explanations for learning a new topic. Presets that can be mixed would fit that better than a single locked mode.

What you can do now

Because the templates are only a proposal, there is nothing official to switch on. But the underlying goal is reachable today with custom instructions. Our briefing suggested saying who you are, who you write for and how blunt the answers should be.

Short and specific usually works better than vague. A line such as "plain language, no flattery, flag anything you are unsure of" tells the assistant far more than "be helpful."

If you want to test the idea, write one instruction set for a week and compare the answers with your usual ones. That gives you a personal template of your own, and a clearer view of whether an official menu would have helped you.

What to watch

Try it

Open ChatGPT's custom instructions and write two or three lines on who you are, who you write for and how blunt you want answers. Use it for a week, then compare.

Sources and links

Liked it? Say why

Add a line for other readers.

No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.

Listen to the AI Wire headlines

Listen to the AI Wire

Listen to our daily AI shows every morning: on your commute, in the shower, wherever you are

Listen to the daily shows

First reported by community.openai.com · Read the original. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.

More AI news