Slide 1
One of the things most insightful people seem to do is seek out other interesting people and have amazing conversations with them. You might have yourself noticed that almost any good idea came in conversation with someone else, or was stoked by it.
Why keep all these insights to yourself?
What if you could scale conversations?
And I know what you're thinking. "You can't just take a private, high-context conversation and publish it. I don't know if I'd event want that"
And if you're thinking we'll solve all those problems with AI, you'd be... WRONG.
Slide 2
We're interested in what happens when humans and AI can work together at this problem. AI handles the logistics of publication, guided minimally by humans at conversation-time.
Slide 3
A three line explainer would be
- A tool for scaling insight in ordinary conversations
- AI sucks at generating insight, but pretty good at structure, polish, and contextualize
- In flowing conversation, occasionally drop "wishes" for the final polished version, to guide AI
This is not about "AI meeting notes". This is about wide dissemination. This is about making everybody a blogger, to really scale.
But not just on the production end. you don't publish a single "final" blogpost; you distribute a raw "blogprompt" of transcript+wishes, that can be "recompiled" for each individual consumer, based on their context. This is not a single publication, but a dynamic, lively republication. Hence "repub.live"
Slide 4
Each of the challenges:
can be addressed with cooperative "wishing" by humans and "wish-fulfilling" by the AI. Individual, semi-formal wishes can be quickly made without having to specify the details.
These aren't like command-line commands, or even AI voice commands. You minimally specify. Hence "wish".
Here's an example, for privacy: