I first saw this quote in @nntaleb's book “Dynamic Hedging” which I consider the best book for financial risk management in the real world. https://t.co/SqDEzWZLEE pic.twitter.com/N3LvmiD0X7
— hardmaru (@hardmaru) April 19, 2019
I first saw this quote in @nntaleb's book “Dynamic Hedging” which I consider the best book for financial risk management in the real world. https://t.co/SqDEzWZLEE pic.twitter.com/N3LvmiD0X7
— hardmaru (@hardmaru) April 19, 2019
When ML and Data Science are the death of a good company: A cautionary tale
— hardmaru (@hardmaru) April 19, 2019
“All of them were very bright people, they could write great Medium blog posts and give inspiring TED talks, but collectively they had very little real world industry experience.” https://t.co/vUAJLEDMxI
"A data scientist is someone who knows more statistics than any software engineer, and more about how to maintain a website without having to schedule two days of downtime than any statistician."
— Joel Grus ☕ (@joelgrus) April 18, 2019
I just published "Data as Prior/Innate knowledge for Deep Learning models" https://t.co/IdvpV0UzTE - would love to get feedback since it is both a simple idea as well as a bit of a stretch
— Xavier (@xamat) April 18, 2019
Since this is going relatively viral, let me recommend an episode of my favorite podcast @CitationsPod https://t.co/VmkHtafIRU
— tj mahr 🍕🍍 (@tjmahr) April 17, 2019
I'm reminded of when Zuckerberg told Congress he wanted to do better on privacy during the same week that Facebook lobbyists were working to gut Illinois privacy laws https://t.co/bJZj7ROvx0 pic.twitter.com/IcZ9yCp4la
— Rachel Thomas (@math_rachel) April 17, 2019
Nice article on AI risks from an international strategic perspective by the always insightful @paul_scharre I strongly support his call for governments/militaries to initiate AI safety conversations. https://t.co/qZH9QVON7Z
— Thomas G. Dietterich (@tdietterich) April 17, 2019
🙌 Another great piece by @rdpeng…
— Mara Averick (@dataandme) April 17, 2019
"Tukey, Design Thinking, and Better Questions" https://t.co/Ql83tJPZSI via @simplystats pic.twitter.com/NOBzXuFpE0
“Behind the scenes, Kaggle has already done the hard job of finding a company with data and a problem, translating a problem into a prediction task, figuring out which data can be used, possibly doing some data cleanup and merging, deciding on a performance metric, etc...“ https://t.co/Y3XqkeieJ6
— hardmaru (@hardmaru) April 17, 2019
A crop of young hardware startups are about to arrive and AMD are awfully close to "close enough" with ROCm CUDA compatibility. Frameworks are abstracted enough that CUDA might not be a moat for that much longer - small nudges like this in the ecosystem may be deciders.
— Smerity (@Smerity) April 17, 2019
NVIDIA claim they prohibited data center use as "GeForce and TITAN GPUs were never designed for data center deployments" but their EULA notes "except that blockchain processing in a datacenter is permitted" ...
— Smerity (@Smerity) April 17, 2019
Lucky blockchain is somehow not magically impacted by these issues 🤔
What's the best way to approach an interview? As one doctor learned, authenticity offered the ideal path to success. This week's #SciMagWorkingLife. https://t.co/KghKYoPpwA
— Science Magazine (@sciencemagazine) April 16, 2019