Chapter
The Future of Intelligent Agents and Their Ability to Learn.
29:54 - 40:39 (10:45)
Gradient-based model predictive control can be used to help intelligent agents do planning, reasoning, and learning, allowing agents to interact with complicated physical systems and environments.
Clips
This podcast discusses model predictive control (MPC) and highlights how MPC uses a predictive model of a system to make decisions on actions to be taken.
29:54 - 32:49 (02:55)
Summary
This podcast discusses model predictive control (MPC) and highlights how MPC uses a predictive model of a system to make decisions on actions to be taken. The podcast explores how MPC can be implemented through deep learning architectures to generate real-time control plans, and how it differs from classical optimal control.
ChapterThe Future of Intelligent Agents and Their Ability to Learn.
Episode#258 – Yann LeCun: Dark Matter of Intelligence and Self-Supervised Learning
PodcastLex Fridman Podcast
The challenge for the next decade of AI is to get machines to learn predictive models that can deal with uncertainty and real-world complexity such as physical systems, people, and behavior.
32:49 - 36:47 (03:57)
Summary
The challenge for the next decade of AI is to get machines to learn predictive models that can deal with uncertainty and real-world complexity such as physical systems, people, and behavior.
ChapterThe Future of Intelligent Agents and Their Ability to Learn.
Episode#258 – Yann LeCun: Dark Matter of Intelligence and Self-Supervised Learning
PodcastLex Fridman Podcast
The use of gradient-based optimization in intelligent agents allows them to make reasoned decisions by predicting objective functions, leading to a more efficient approach than zero-th order, gradient-free optimization.
36:47 - 40:39 (03:52)
Summary
The use of gradient-based optimization in intelligent agents allows them to make reasoned decisions by predicting objective functions, leading to a more efficient approach than zero-th order, gradient-free optimization.