I’d like to suggest a broader direction for future versions of ChatGPT: giving users a more natural way to discuss audiovisual and cultural material with AI.
Today, users can often provide text about a video, interview, lecture, literary work, or cultural recording and ask ChatGPT to analyze it. But I think there is a larger opportunity: allowing AI to understand the original source material more directly, rather than relying entirely on information manually extracted and provided by the user.
For example, a user might want to discuss a recording of a writer, poet, philosopher, historian, musician, or lecturer speaking in their native language. Ideally, the AI should be able to understand not only the literal words, but also linguistic nuance, context, cultural references, tone, ideas, and other meaningful aspects of the original material.
I think this could be especially valuable for multilingual users and for languages with relatively limited digital resources.
For example, Georgian is spoken by only a few million people, but it has a rich literary, philosophical, historical, and artistic tradition. I believe AI should not measure the value of a language by the number of people who speak it.
More broadly, this could help move ChatGPT from understanding only what a user tells it about something toward understanding the original information the user is actually referring to.
This could open interesting possibilities for lectures, interviews, documentaries, literary readings, historical recordings, music, theatre, and many other forms of human expression.
I’m interested in whether OpenAI is already exploring this broader direction, and whether other users see value in it as well.
I’d be happy to elaborate further and share a concrete use case if the idea is interesting to the community or the OpenAI team.