Jan 28, 2023

Amazing New Google AI Learns Millions of Tasks In Minutes w/o Training + Reinforcement Learning

Posted by in categories: innovation, robotics/AI

Deep Learning AI Specialization:
Breakthrough Google DeepMind Adaptive AI agents, known as AdA, are able to learning millions of tasks within minutes each with reinforcement learning, matching the speed and ability of a skilled human gamer, all without using training data. A New InstructPix2Pix text to image artificial intelligence can do photo editing with simple text instructions. A new dataset called OMMO enables novel view synthesis, surface reconstruction, and multi-modal NeRF.

AI News Timestamps:
0:00 New Google DeepMind Reinforcement Learning AI
3:11 New InstructPix2Pix Artificial Intelligence.
6:29 OMMO NeRF View Synthesis.

#technology #tech #ai

Leave a reply