Chapter
The Future of Deep Learning and Innovation in Hardware
The efficiency of deep learning for specific tasks is driving innovation in hardware, which could lead to discovering intelligent systems that operate differently than deep learning. By using different insights, such as making it more dynamic or incentivizing, and adding dendrites to neurons, it is possible to solve problems in continuous and rapid learning.
Clips
Researchers observed birds and discovered their wing-twisting secret for better turning.
2:03:35 - 2:06:12 (02:37)
Summary
Researchers observed birds and discovered their wing-twisting secret for better turning. As for the Neo Cortex in machines, there may be decades of innovation to come for building flying machines that surpass birds in their capabilities.
ChapterThe Future of Deep Learning and Innovation in Hardware
Episode#208 – Jeff Hawkins: The Thousand Brains Theory of Intelligence
PodcastLex Fridman Podcast
The effectiveness of deep learning for specific tasks is driving innovation in hardware, which may lead to the discovery of intelligence systems that operate differently or on a larger scale than deep learning.
2:06:12 - 2:10:05 (03:53)
Summary
The effectiveness of deep learning for specific tasks is driving innovation in hardware, which may lead to the discovery of intelligence systems that operate differently or on a larger scale than deep learning. However, adding dendrites to neural models can solve problems in continuous and rapid learning similar to biological systems.