Hi HN! My friend and I (former MIT Media Lab researchers) started a project that has gotten a lot of momentum, and wanted to see what you all think. It’s a Jarvis-like AI that uses HDMI pass-through to control and overlay assistive content on your TV.
Basically, we're trying take the hassle out the things you do every day: In the morning, Veil is ready to call you a Lyft, show your commute time, and summarize today’s agenda. In the evening, Veil’s ready with your favorite Spotify playlist and Netflix shows, all without asking or grabbing a remote control/sifting through apps.
We built this because Amazon Echo & Google Home were a bit of a let-down. They’re not really “assistants” so much as servants; they can’t do anything without being told, and don’t know who’s using them.
Our AI-device aims to be:
1) Proactive - it anticipates needs and shows information on the TV without being asked. A speaker can never to this without blurting out. That would get annoying.
2) User-aware - it’s always helpful to the current user (or group of users) and accesses their accounts automatically (current AIs are tied to one account).
3) Always-ready - our device sits BETWEEN your TV and existing media boxes. It overlays content on every input, so you never have to change input to use it. It's always ready to help and even manages input switching (which is a terrible experience for most people).
We're using Google's Voice/Language API and eventually hope to integrate Assistant.
Would love to hear your thoughts!
If you’re interested staying in touch with our development, or getting your hands on our alpha unit, check our project page or PM me!
Basically, we're trying take the hassle out the things you do every day: In the morning, Veil is ready to call you a Lyft, show your commute time, and summarize today’s agenda. In the evening, Veil’s ready with your favorite Spotify playlist and Netflix shows, all without asking or grabbing a remote control/sifting through apps.
http://diatom.tv/veil
We built this because Amazon Echo & Google Home were a bit of a let-down. They’re not really “assistants” so much as servants; they can’t do anything without being told, and don’t know who’s using them.
Our AI-device aims to be:
1) Proactive - it anticipates needs and shows information on the TV without being asked. A speaker can never to this without blurting out. That would get annoying.
2) User-aware - it’s always helpful to the current user (or group of users) and accesses their accounts automatically (current AIs are tied to one account).
3) Always-ready - our device sits BETWEEN your TV and existing media boxes. It overlays content on every input, so you never have to change input to use it. It's always ready to help and even manages input switching (which is a terrible experience for most people).
We're using Google's Voice/Language API and eventually hope to integrate Assistant.
Would love to hear your thoughts! If you’re interested staying in touch with our development, or getting your hands on our alpha unit, check our project page or PM me!