Groundbreaking AI advancements, from GPT-5.1 and a novel text-to-speech model to AI agents mastering video games and a robot army. Witness AI's enhanced visual perception, mirroring human understanding.
https://deepmind.google/blog/teaching-ai-to-see-the-world-more-like-we-do
https://time-to-move.github.io
https://github.com/WeiboAI/VibeThinker
https://stepaudiollm.github.io/step-audio-editx
https://github.com/360CVGroup/EVTAR
https://openai.com/index/gpt-5-1
https://www.worldlabs.ai/blog/marble-world-model
https://pointscoder.github.io/PhysWorld_Web
00:00 - AI news intro
00:51 - Vision alignment
04:29 - Time to Move
06:22 - VibeThinker 1.5B
08:28 - Step Audio EditX
13:39 - EVTAR
15:33 - GPT 5.1
18:39 - AI persona marriage
19:58 - Artlist
21:18 - Ubtech robot army
22:20 - Unitree doing chores
23:24 - Marble world model
28:00 - Lumine
30:47 - SIMA2
34:57 - Ernie 5.0
40:13 - PhysWorld