Turn user corrections into verified shared knowledge

ChatGPT Feature requests
1

I would like to suggest a mechanism that allows useful factual corrections from users to benefit other ChatGPT users — but only after independent verification.

The principle would be:

User correction → independent verification by ChatGPT → validation from reliable public sources → reusable knowledge for future users.

A concrete example:

I asked ChatGPT where I could park a motorcycle near 13 Rue Saint-Rémi in Bordeaux, France. ChatGPT searched for nearby motorcycle parking but failed to identify the motorcycle parking area at Place Jean-Jaurès.

I told ChatGPT that this parking area exists.

At that point, my statement was merely user-provided information and should obviously not be trusted automatically.

However, ChatGPT was then able to independently verify the information using public sources.

Once independently verified, the information no longer depends on my claim. Yet the correction remains confined to my conversation and does not directly improve the answer that another user might receive tomorrow.

I think there is an opportunity to create a carefully controlled system where:

  1. A user identifies a potentially incorrect or missing factual piece of information.
  2. ChatGPT searches independent, reliable sources to verify it.
  3. If the information can be sufficiently verified, it becomes a candidate for a shared factual knowledge layer.
  4. Future users can benefit from that verified information without their answers relying on the original user’s statement.
  5. Information that changes over time could retain its sources and verification date so that it can be rechecked when necessary.

This would not mean allowing users to directly modify ChatGPT’s knowledge. The crucial distinction is that the user only provides the lead; independent sources provide the evidence.

This could be particularly valuable for local and frequently changing information — parking locations, accessibility information, public facilities, transportation details, local regulations, business information, etc. — where individual users often notice omissions that a general search initially misses.

In other words, ChatGPT could turn certain user corrections into verified, sourced and reusable knowledge, rather than allowing a useful discovery to disappear when the conversation ends.