1. Voice is a great interface to communicate what's in your mind to a computer and voice agents in general, are very good at holding a conversation but having a voice to voice conversation about why your system broke down at 2 am and you lost a customer, doesn’t work.
2. Are good at generating UI elements, but there, you write something in the chat window, it generates an artifact, and then you go move to and fro from the chat window to the artifact. That's also not a great experience.
3. Agents are great at having holistic context from multiple apps but chatting with them again makes you think linearly by scrolling up and down
So, we’re building a multimodal JARVIS like interaction harness that is spatial and allows you to express your thoughts interact with the agent in a way that’s natural to you and let’s agent explain it’s work or whatever you asked of it in a way it can be consumed easily by your brain.
Today’s AI tools landscape -
1. Voice is a great interface to communicate what's in your mind to a computer and voice agents in general, are very good at holding a conversation but having a voice to voice conversation about why your system broke down at 2 am and you lost a customer, doesn’t work.
2. Are good at generating UI elements, but there, you write something in the chat window, it generates an artifact, and then you go move to and fro from the chat window to the artifact. That's also not a great experience.
3. Agents are great at having holistic context from multiple apps but chatting with them again makes you think linearly by scrolling up and down
So, we’re building a multimodal JARVIS like interaction harness that is spatial and allows you to express your thoughts interact with the agent in a way that’s natural to you and let’s agent explain it’s work or whatever you asked of it in a way it can be consumed easily by your brain.
Looks great! I've been annoyed at how useless codex's voice mode is in being an actual collaborator