Week 4 of the #WFH experiment. With the uptick in remote collab and communication, staying on top of emails is critical. It’s important to get help wherever you can.
Could the future of machine learning computation be... SQL query engines? Jankov et al. show the power of distributed databases for large scale machine learning. https://t.co/DBOsV7V2cr
Not a marine biologist, but I played one on TV! @CHANGINGSEAS produced a great documentary about Cordell Bank National Marine Sanctuary which aired on PBS last night - including BAUE’s ongoing work with NOAA to document and conserve. https://t.co/ljpfjNM42Q @GUEhq
According to IDC, Pure grew faster than any of the top 5 vendors at 22% year over year in 2019 Q1. https://t.co/0jb9EjQRKs #ThankYouCustomers#IDCTrackerQ12019
Things you don’t want to hear on your plane “one of the flight computers is broken - but we have 2, so maintenance is just going to disable one and we’ll be on our way”. 🙄
As programmers grow in experience, they develop their own unique styles. The stylistic fingerprints survive even compilation to binary executables with optimisations and basic obfuscation enabled! It's hard to be anonymous as a programmer... https://t.co/xVS4KWAI7T
Interesting paper from Google that @svitlanacs shared highlighting the relative (and hidden) complexity in ML systems: https://t.co/rs8fmvIknm
tl;dr; the actual ML is dwarfed by everything around it managing data
It takes a human being about 14 thousand brain-hours to learn to run. An AI can learn to do it in less than half as many CPU-hours, but the results are like this.
Storing off my weekend VR project, managed to brick my consumer HDD NAS device into "Emergency Mode" with partition table corruption. Paying for the sins of my past I guess - looking for a good data recovery outfit. Should have been using a #FlashBlade 😀