Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I think part of the problem is not all conversations are obviously 'short and shallow' or 'long and broad'. What starts out casual might unearth something that needs going into deeper, and going in deep on something might not have results worth documenting yet.

To me this points out a gap in our software for knowledge archival. The closest concept Slack has is "pinning" a message to a channel, which has very limited use for this. Really what you want is to be able to have a conversation with someone in any manner (casual or formal meeting) and be able to easily extract the conclusions and decisions made into a different location, ideally through an easily discoverable UI in the software already being used.

In an ideal world my chat software has knowledge of the projects I'm working on (through some protocol to something else that acts as a source of truth) and I can select a message/file upload in chat and tell it to archive it against a project. It can easily fill in the details for me about who made this decision and when. Ability to effortlessly organise information arising from a connversation should a priority I think, this might remove the need for having seperate software depending on how "deep" you think a conversation is going to be.



I know that AI is a pile of buzzwords, but.... is this something we could train AI to do reasonably well? A flow like "click button and highlight conversation start/end" and it picks out the relevant messages and tries to generate a set of takeaways? (and allows user to edit them before saving).


AI (or some primitive concept of...) could be useful in some aspect, perhaps for sugggesting tags or categories the information looks like it could belong to. Actually extracting the "facts and takeaways" sounds very hard but then I'm not a linguist so maybe there are methods.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: