Bodi Yuan

Cofounder of Reverie AI | Berkeley PhD without dropout. Turning videos into playable worlds.

[Bodi Yuan @BodiYuan]

@drfeifei @smallfly @theworldlabs Clapping out loud for this, Hugues! This is soooo cool! How much time to build this?

[Bodi Yuan @BodiYuan]

@VivianCaiIAm @UseCorgi @mercury @AgentWebPro @RespanAI @arceuslegal @JoinPond wow nice!

[Bodi Yuan @BodiYuan]

@measure_plan wow this is sooo cool!

[Bodi Yuan @BodiYuan]

@peggy_wang Congrats Peggy!

[Bodi Yuan @BodiYuan]

@FishAudio Ah nice!!! We need it desperately!

[Bodi Yuan @BodiYuan]

@ionleu https://t.co/VaTvflunzt

[Bodi Yuan @BodiYuan]

The next AI interface is not a text box. Announcing Reverie Interaction Model R v0.2.

Humans interact with the world across all modalities: we speak, listen, read, look, touch, gesture, and react. AI should be able to do the same. Model R v0.2 gives AI a richer way to understand the world: not only reading and listening, but seeing, even sensing touch and actions, and responding across modalities including text, audio, and video. This is an all-to-all multimodal interface.

For interaction to feel natural, it has to happen in real time. All-to-all multimodal interaction is already hard; making it instant and fluid is even harder. But that is what turns AI from something you query into something you interact with.

The text box was the first great AI interface, but it will not be the only one, especially for entertainment. People do not just want to type instructions. They want to watch, touch, play, shape, and participate.

More coming soon.

[Bodi Yuan @BodiYuan]

@lightsilver323 https://t.co/VaTvflunzt

[Bodi Yuan @BodiYuan]

The next AI interface is not a text box. Announcing Reverie Interaction Model R v0.2.

Humans interact with the world across all modalities: we speak, listen, read, look, touch, gesture, and react. AI should be able to do the same. Model R v0.2 gives AI a richer way to understand the world: not only reading and listening, but seeing, even sensing touch and actions, and responding across modalities including text, audio, and video. This is an all-to-all multimodal interface.

For interaction to feel natural, it has to happen in real time. All-to-all multimodal interaction is already hard; making it instant and fluid is even harder. But that is what turns AI from something you query into something you interact with.

The text box was the first great AI interface, but it will not be the only one, especially for entertainment. People do not just want to type instructions. They want to watch, touch, play, shape, and participate.

More coming soon.

[Bodi Yuan @BodiYuan]

@EmilyJarvis2021 🤪

[Bodi Yuan @BodiYuan]

The next AI interface is not a text box. Announcing Reverie Interaction Model R v0.2.

Humans interact with the world across all modalities: we speak, listen, read, look, touch, gesture, and react. AI should be able to do the same. Model R v0.2 gives AI a richer way to understand the world: not only reading and listening, but seeing, even sensing touch and actions, and responding across modalities including text, audio, and video. This is an all-to-all multimodal interface.

For interaction to feel natural, it has to happen in real time. All-to-all multimodal interaction is already hard; making it instant and fluid is even harder. But that is what turns AI from something you query into something you interact with.

The text box was the first great AI interface, but it will not be the only one, especially for entertainment. People do not just want to type instructions. They want to watch, touch, play, shape, and participate.

More coming soon.

[Bodi Yuan @BodiYuan]

@Robnin thank you!

[Bodi Yuan @BodiYuan]

@heather129770 will release our api soon!

[Bodi Yuan @BodiYuan]

🫥

[Justine Moore @venturetwins]

🚨 New market map alert!

Our team @a16z consumer dove into the first killer use case of AI: making creative content.

A recap of the companies tackling generation and editing + what we're hoping to see next 👇