What Serverless Computing Should Become π must-read paper from @ucbrise outlines current state of serverless computing & lists research challenges in five areas: abstractions, systems, networking, security, and architecture https://t.co/4l18g9WCht pic.twitter.com/tW8rLLFVq2
β Ben Lorica η½ηε‘ (@bigdata) February 10, 2019
This new study still is limited to testing recall 30 mins later, and without internet access. So it's still missing the value of looking up unfamiliar concepts, creating links, collaborative note-taking...
β Jeremy Howard (@jeremyphoward) February 10, 2019
(plus of course many folks can't comfortably write by hand) https://t.co/CTd4mOs4gX
Counterfactual universe @fhuszar has some choice words for your latest bit of alchemy. pic.twitter.com/xJxvJJ8FM6
β Kaixhin (@KaiLashArul) February 10, 2019
Here's some speculation on Tesla AP V9 using inception stem https://t.co/Sf7LuVvazy
β shreyas gite (@shreyasgite) February 10, 2019
Why don't modern architectures use the Inception stem? Did people try it and find it doesn't help over 3 3x3 convs? Or has it been forgotten?
β Jeremy Howard (@jeremyphoward) February 10, 2019
It seems like a smart design to me, but I haven't done experiments...
This Pixar short is awesome. Really captures the negative dynamic in an office setting that doesn't promote inclusion, and how much better things can be when we all make the effort to be inclusive. https://t.co/DGCgaSRO0d
β Jeff Dean (@JeffDean) February 10, 2019
I'm quoted in this @WIRED article on Twitter malicious applications and the research by @zubair_shafiq @shehrozeFarooqi on how to detect themhttps://t.co/egf8ZSkGpU
β Gianluca Stringhini (@gianluca_string) February 9, 2019
The UN-surprising truth about building a machine learning product. However, it has a missing part: convincing the client to invest in improved data collection process. https://t.co/BICJw6unUS
β Leonid Boytsov (@srchvrs) February 9, 2019
I saw this floating around in my LinkedIn. I think it's the best advice I would give my 10 years younger self. pic.twitter.com/vgdqD3DYv6
β Delip Rao (@deliprao) February 8, 2019
Carol Dweck on misunderstandings about growth mindset:
β Rachel Thomas (@math_rachel) February 8, 2019
A lot of teachers say praise the effort, not the outcome. I say thatβs wrong...Itβs not just effort, but strategy. You donβt want students redoubling their efforts with the same ineffective strategies.https://t.co/pymtEX0hvx pic.twitter.com/HsOFHmIHb3
It seems like we are going back to the UNIX-era of computing where researchers are allocated time slices for compute. If you are at a university lab running large cloud-based compute jobs, do you use something off the shelf, or does your lab try to write its own management tools?
β hardmaru (@hardmaru) February 7, 2019
Much commentary on AI in journalism is about AI displacing jobs. But natural language generation tech can only go so far today.
β Hilary Mason (@hmason) February 6, 2019
I think we're likely to see these tools put to work by reporters, not instead of them. https://t.co/aGLDog0Mtd
