Chapter
The Nature of Learning and Deep Learning Interpretation
23:06 - 28:58 (05:52)
In this podcast, the speaker questions the nature of learning and the interpretations behind deep learning. They argue that the optimal solution for learning lies in shadow networks, rather than deep learning.
Clips
The speaker criticizes the use of deep learning and neural networks in machine learning, likening it to fantasy and emphasizing the importance of incorporating mathematical knowledge in the field.
23:06 - 25:01 (01:55)
Deep Learning
Summary
The speaker criticizes the use of deep learning and neural networks in machine learning, likening it to fantasy and emphasizing the importance of incorporating mathematical knowledge in the field.
ChapterThe Nature of Learning and Deep Learning Interpretation
EpisodeVladimir Vapnik: Statistical Learning
PodcastLex Fridman Podcast
The optimal solution for learning lies in the shadow network, rather than deep learning, as deep learning is a learning system that is not yet mathematically understood.
25:01 - 28:58 (03:56)
Deep Learning
Summary
The optimal solution for learning lies in the shadow network, rather than deep learning, as deep learning is a learning system that is not yet mathematically understood. Despite this, deep learning has beaten the best human player in learning tasks.