Karpathy's tweet from 2017 about gradient descent
Just came across this post containing a screenshot of Andrej Karpathy humorously tweeting in 2017 about Gradient Descent being able to code better than humans. Now that the AI models have good coding capability, Karpathy's tweet has aged well.
By the way, I did not know what Gradient Descent is, so I had to look it up.
Gradient descent is a method for unconstrained mathematical optimization. It is a first-order iterative algorithm for minimizing a differentiable multivariate function.
This is the 1-line explanation, but I need to do more reading about this. Gotta go.
Webmentions