A ChatGPT user asks for a 'Tell Sam' button to send ideas straight to OpenAI

A button with a friendly name

A post on OpenAI's community forum, as community.openai.com shows, makes a simple request. It asks for a small feature called "Tell Sam", an icon inside ChatGPT that lets users send ideas, suggestions, frustrations and product improvements directly to OpenAI.

The name is a nod to MySpace. The author recalls that when you joined that social network, its founder Tom Anderson automatically became your first friend. Millions of people felt a small personal link to the person running the platform. The post asks whether ChatGPT could offer something similar, with a link to Sam rather than Tom.

This is a user's proposal. It is not an OpenAI announcement, and the post gives no sign that the company has agreed to anything.

How the idea would work

The post sketches a scene. You are chatting with ChatGPT and an interesting product idea comes up. You remark that OpenAI should consider it, and ChatGPT asks, "Would you like to Tell Sam?"

One click would have ChatGPT summarise your suggestion. You could review and edit it before it goes to a central feedback system. You might also get an acknowledgement, and perhaps see when your idea fed into a wider product discussion.

The post then describes what AI could do with the pile of submissions:

• Read and sort them into categories.
• Find frustrations and unmet needs that keep coming up.
• Spot genuinely new ideas that might otherwise be missed.
• Group similar proposals and measure how much demand each has.
• Write regular summaries for product teams and leadership.

The author adds that nobody should pretend Sam Altman personally reads every suggestion. The point is that something capable of weighing the ideas would.

Why a doctor wants this

The author describes himself as an Australian general practitioner. He says ChatGPT has greatly increased his personal agency over the past year. He has used it to write submissions and suggestions to professional organisations, government departments and businesses.

He is realistic about the results. He says 99% of these ideas will probably never be implemented, and he is fine with that. His argument is that there is real psychological value in having an idea, putting it into words and knowing that someone, or something able to judge it sensibly, has listened.

He also argues that OpenAI would gain. A million people asking for the same feature would show clear demand, and one person with an unexpected idea might point to a new opportunity. A well-built system, he says, could tell the two apart.

The bigger claim, and its limits

The post ends with a larger vision. It suggests the approach could become a model for AI systems that help governments, institutions and organisations listen to millions of individuals. The author calls "Tell Sam" a small interface feature with potentially enormous implications for engagement, product development and human agency.

That is a big claim for a short forum post, and the post does not back it up. It does not say how submissions would be ranked or how privacy would be protected beyond saying it should be. It does not explain how people would know their ideas were handled fairly. The author states those principles (voluntary, privacy-protected, transparent) but does not design them.

It is also one voice. We cannot say how many other users share the view, and we have seen no reply from OpenAI. The post is best read as a sign of mood: people who use these tools heavily want to feel heard by the company that makes them.