Topic
Jeremy Howard
In this fireside chat at PyTorchCon, Jeremy Howard reflects on his early days as a contrarian who championed PyTorch over Google's TensorFlow and predicted the rise of large language models before they were mainstream. He recounts how he and Fast AI initially faced skepticism for investing in an "ob …
Lex Fridman
Jeremy Howard, founder of fast.ai and a distinguished research scientist at the University of San Francisco, emphasizes making deep learning accessible through practical application rather than theoretical obscurity. He traces his interest in programming back to high school projects on a Commodore 6 …
Lex Fridman
Jeremy Howard opens his discussion with a stark critique of the current state of deep learning research, asserting that most academic efforts in this field are essentially a waste of time. He argues that the scientific publishing system forces researchers to focus on topics their peers already under …
Lex Fridman
In this discussion, Jeremy Howard highlights a significant breakthrough in neural network training known as super convergence, originally discovered by researcher Leslie Smith. This phenomenon reveals that certain networks with specific high-parameter settings can be trained ten times faster and ach …