Today we’re sharing our AI principles and practices. How AI is developed and used will have a significant impact on society for many years to come. We feel a deep responsibility to get this right. blog.google/topics/ai/ai-p…

most common neural net mistakes: 1) you didn't try to overfit a single batch first. 2) you forgot to toggle train/eval mode for the net. 3) you forgot to .zero_grad() (in pytorch) before .backward(). 4) you passed softmaxed outputs to a loss that expects raw logits. ; others? :)

Can we predict side effects of drug combinations? Our paper gains insights on how different drug combinations interacted with protein networks to create side effects -- before they happen. academic.oup.com/bioinformatics… news.stanford.edu/2018/07/10/ai-…

IBM Watson reportedly recommended cancer treatments that were "unsafe and incorrect" gizmo.do/SR1so9O

Chatted about my journey through AI, including stories from early days of Google Brain & Coursera, plus Landing AI and deeplearning.ai. Hope you'll find useful ideas and inspiration from this. Thanks Kevin Scott for doing this! behindthetech.libsynpro.com/003-andrew-ng-…

Deep learning needs more theory. Many workshops I have helped organize in the last decade (e.g. at IPAM) have attempted to bring ML/DL folks together with mathematicians and... quantamagazine.org/foundations-bu…

Should all layers be initialized the same, or is it important to change the initialization for different types of layers (e.g. LSTM weights vs CNN weights)? Read the new AI Notes tutorial on initialization and let us know your thoughts in the comments: bit.ly/2XKKyu3

Early detection of sepsis utilizing deep learning on electronic health record event sequences - ideas from my Master's thesis enhanced by health scientists arxiv.org/abs/1906.02956

Introducing The Batch, a new weekly newsletter from @deeplearningai_! AI is growing so fast that it’s hard to know what news is useful or valuable. Subscribe to The Batch for a curated report on the latest AI research and industry-shaping events: bit.ly/2YYzp9b

TensorFlow 2.0.0 release is out now: github.com/tensorflow/ten… Huge milestone for the machine learning ecosystem. Congratulations to the hundreds (yes) of people who have contributed!

CookGAN: Meal Image Synthesis from Ingredients pdf: arxiv.org/pdf/2002.11493… abs: arxiv.org/abs/2002.11493

Taking a bathroom break from #COVID19 to go to the smart toilet, which uses deep learning #AI to analyze all of one's excrements 💩 nature.com/articles/s4155… Nature Biomedical Engineering Stanford Medicine Stanford Engineering

I just published TFServingCache: A Distributed LRU-Cache for Serving TensorFlow Models in Kubernetes link.medium.com/W9UvrtrEBjb #TensorFlow #MachineLearning

Digging Into My Year of Spotify Streaming with Track Embeddings by Mads Kalør buff.ly/3fomFmG

Read this thread. x.com/fchollet/statu…

Read my interview with WIRED on artificial intelligence : wired.com/story/emmanuel…

Get a free draft copy of my book on how to structure Machine Learning projects: mlyearning.org I’d started this before but got distracted building Deep Learning Specialization; I’m now rebooting this. Sign up to get free chapters as they’re released!

Every paragraph of this terrific Palantir story is absolutely bonkers. bloomberg.com/features/2018-…

massive, AI-powered roach breeding farm produces more than six billion cockroaches per year gizmodo.com/this-ai-contro… << what could possibly go wrong?

Nice article about a neat trick! A fast sparse multiplication into an embedding is indeed really useful in terms of performance and effectiveness for many tasks involving categorical input features x.com/math_rachel/st…