Hear Dr. @ChrisBishopMSFT talk about the past, present and future of AI research on the #MSRPodcast. He also discusses the modern view of machine learning as Director of @MSFTResearchCam: https://t.co/XlerNeAbjP #NeurIPS2018
— Microsoft Research (@MSFTResearch) November 28, 2018
I swear this 📄 should be universal required reading! 🏆💖 to @kwbroman and @kara_woo for 💸 to make it open access ⇨ #NoExcuses
— Mara Averick (@dataandme) November 28, 2018
"Data Organization in Spreadsheets" https://t.co/XBoXNE5fOK
One common objection to sharing data from a published study: “what if someone finds a mistake in my data/analysis?” 🤦♂️
— Ben Hamner (@benhamner) November 28, 2018
Political Scientists looking for contemporary collective action problems to study should take a look at this serious open-source kerfuffle. It's got everything: perverse incentives, signaling games, free riders, negative externalities, and in- and out-group dynamics! https://t.co/zWk8lLGzN0
— Drew Conway (@drewconway) November 27, 2018
Very happy to see this - exactly the kind of thing we hoped fastai would enable. A dramatic improvement in state of the art for damage level classifications after disasters.
— Jeremy Howard (@jeremyphoward) November 27, 2018
Great job @SanthoshShetty! https://t.co/coHYXZYOv9
How cheap labor labeling training data in "data factories" drives China’s A.I. ambitions https://t.co/cWwhN40lB5
— Peter Skomoroch (@peteskomoroch) November 27, 2018
It shows again how much work open source maintenance can be, if your library is successful you have a ton of responsibilities and can cause severe damage
— Felix Krause (@KrauseFx) November 26, 2018
“Researchers who are mobile get more citations and build broader teams of collaborators than those who aren’t, study finds.”
— Alessandro Vespignani (@alexvespi) November 26, 2018
https://t.co/5RBZYb7Haw
"In AI, we have Milgram's ultimate authority figure." -- Peter Haas
— Rachel Thomas (@math_rachel) November 25, 2018
(The Milgram experiment showed that people will obey an authority figure even when it entails harming others. People may be even more likely to trust what an algorithm says.)https://t.co/EpDrERm1H9
US job applicants are getting a single number that determines if they get a job. That number comes from a bollocks AI system that claims to rate their social media activity. And people think 'social credit scores' are just in China 🤔 pic.twitter.com/LbfHG5kQbQ
— Kate Crawford (@katecrawford) November 24, 2018
0. Thinking this time it will be easy. https://t.co/XIRdZWQ1CA
— Chris Albon (@chrisalbon) November 23, 2018
Rough price to train one #bigGAN model if using GCP cloud services. https://t.co/KkVNbXsxUE
— Jack Clark (@jackclarkSF) November 22, 2018
