Things I read last week… (WB 3 Oct 2022)
Google Brain Team’s Imagen Video.
“Improving the efficiency of algorithms for fundamental computations can have a widespread impact, as it can affect the overall speed of a large amount of computations. Matrix multiplication is one such primitive task, occurring in many systems—from neural networks to scientific computing routines. The automatic discovery of algorithms using machine learning offers the prospect of reaching beyond human intuition and outperforming the current best human-designed algorithms.”
– Discovering faster matrix multiplication algorithms with reinforcement learning.
Google’s AudioLM can generates naturalistic sounds from a short snippet.
Google officially launches Equiano subsea cable linking Africa.
Extended the scaling hypothesis (that ML models get better with more computation) to reinforcement learning. [PDF]
The map of engineering. [YouTube]
The US Military’s global transportation system. [YouTube]