DeepMind’s latest AI, called MuZero, man master games while learning how to play, all without being told the rules. Earlier this year, it managed to master dozens of Atari 2600 games from over forty years ago, a great training exercise for the AI because they provide a rich suite of tasks which players must develop sophisticated behavioral strategies to master, and give it an easy progress metric – game score – to optimize against.
Now the MuZero AI learns the rules of the game on its own and models only the most important information, which makes it hyper efficient over standard model-based systems. These skills are all learned using a deep neural network and the only thing needed for MuZero to understand what happens when it takes a certain action and to plan accordingly.
- Pixel Buds Pro use Active Noise Cancellation with Silent Seal to adapt to your ear and help block outside sounds, creating a quiet foundation so your...
- With custom 11 mm speaker drivers and Volume EQ, the earbuds sound amazing at any volume[1]
- Pairs with any Bluetooth 4.0+ device, including Android, iOS, tablets, and laptops[7]
MuZero’s ability to both learn a model of its environment and use it to successfully plan demonstrates a significant advance in reinforcement learning and the pursuit of general purpose algorithms. Its predecessor, AlphaZero, has already been applied to a range of complex problems in chemistry, quantum physics and beyond,” said the team.
