3 ms·
Describing a 3d scene with text is one way using ML to update an NPC, but I feel like the real killer app would be a multi-modal agent. Deepmind recently showed
by arthurdotonce 2y ago
Describing a 3d scene with text is one way using ML to update an NPC, but I feel like the real killer app would be a multi-modal agent. Deepmind recently showed some impressive progress in this sense: https://deepmind.google/discover/blog/sima-generalist-ai-agent-for-3d-virtual-environments/ https://deepmind.google/discover/blog/sima-generalist-ai-age...
- Tantris 2y agoThanks for sharing this link! This would definitely be a silver-bullet, but I'm not aware of such agent having the ability to take and spit out text input! Another example of this idea would be Midjourney's CEO vocal plans to create models able to natively take in 3d input - but no demo yet :)