Entries from May 2017 ↓

• Google now knows when its users go to the store and buy stuff – The Washington Post
• Drop your notepad, reMarkable is coming for your reams
• “Personal kanban”: a time-management system that explodes the myth of multitasking — Quartz
• Leaked regulation: Trump plans to roll back Obamacare birth control mandate – Vox
• Latent semantic analysis – Wikipedia
• data/subreddit-algebra at master · fivethirtyeight/data · GitHub
• Sources: Russians discussed ‘derogatory’ information about Trump and associates during campaign – CNNPolitics.com
• Core Kubernetes: Jazz Improv over Orchestration – Heptio
• Twitter
  RT @kishau: Scathing and so beautifully written. This is just … superb. 👇🏽
• Rebecca Solnit: The Loneliness of Donald Trump | Literary Hub

  A brilliant essay: The most mocked man in the world.
  “Obliviousness is privilege’s form of deprivation”
  http://lithub.com/rebecca-solnit-the-loneliness-of-donald-trump/ h/t @moorehn

• Twitter
  RT @ticiaverveer: A beautiful Etruscan amber carving of a lion’s head.
  500-480 BC #Italy (or lioness?) worn as a pendant…
• Human Resources Isn’t About Humans – Backchannel
• What’s That Bug? – Are we experts yet?
• Twitter
  RT @dabeard: McCain to Australia: #Trump has ‘unsettled’ America; please stick with US & encourage us to help deal with this…
• Windows shattered at Lexington Herald-Leader; suspected bullet damage found | Lexington Herald Leader
• Trump administration plans to minimize civil rights efforts in agencies – The Washington Post
• How Trump Took Hate Groups Mainstream | Mother Jones
• How Congress dismantled federal Internet privacy rules – The Washington Post

  Google funneled money to Republicans in a successful attempt to kill ISP privacy. Not a peep out of Googlers. https://www.washingtonpost.com/politics/how-congress-dismantled-federal-internet-privacy-rules/2017/05/29/7ad06e14-2f5b-11e7-8674-437ddb6e813e_story.html

• First Look at the Essential Phone, Andy Rubin’s Anti-iPhone | WIRED
• contrib.layers.avg_pool2d raises warnings when serializing metagraph · Issue #9939 · tensorflow/tensorflow
• The Bullshitter-in-Chief – Vox

  Trump offers the style of authoritarian propaganda under circumstances of democratic pluralism. https://www.vox.com/policy-and-politics/2017/5/30/15631710/trump-bullshit?utm_campaign=mattyglesias&utm_content=chorus&utm_medium=social&utm_source=twitter https://twitter.com/mattyglesias/status/869538487212670976/photo/1

• [1511.07289] Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)
  We introduce the "exponential linear unit" (ELU) which speeds up learning in deep neural networks and leads to higher classification accuracies. Like rectified linear units (ReLUs), leaky ReLUs (LReLUs) and parametrized ReLUs (PReLUs), ELUs alleviate the vanishing gradient problem via the identity for positive values. However, ELUs have improved learning characteristics compared to the units with other activation functions. In contrast to ReLUs, ELUs have negative values which allows them to push mean unit activations closer to zero like batch normalization but with lower computational complexity. Mean shifts toward zero speed up learning by bringing the normal gradient closer to the unit natural gradient because of a reduced bias shift effect. While LReLUs and PReLUs have negative values, too, they do not ensure a noise-robust deactivation state. ELUs saturate to a negative value with smaller inputs and thereby decrease the forward propagated variation and information. Therefore, ELUs code the degree of presence of particular phenomena in the input, while they do not quantitatively model the degree of their absence. In experiments, ELUs lead not only to faster learning, but also to significantly better generalization performance than ReLUs and LReLUs on networks with more than 5 layers. On CIFAR-100 ELUs networks significantly outperform ReLU networks with batch normalization while batch normalization does not improve ELU networks. ELU networks are among the top 10 reported CIFAR-10 results and yield the best published result on CIFAR-100, without resorting to multi-view evaluation or model averaging. On ImageNet, ELU networks considerably speed up learning compared to a ReLU network with the same architecture, obtaining less than 10% classification error for a single crop, single model network.
• Twitter
  RT @EdJoyce: ‘An enormous loss’ 900 miles of the #GreatBarrierReef have bleached severely since 2016. …
• Twitter
  RT @ColinKahl: In the wake of Trump’s trip, our closest allies now see the United States as a threat to European peace.…
• Twitter
  RT @TVMaury: When the Nordic prime ministers are openly mocking #PresidentTrump and the #Saudis ..
• Twitter
  RT @EdJoyce: The great #Greenland meltdown: "Nobody expected the ice sheet to lose so much mass so quickly" @uciess…
• Twitter
  RT @GarettJones: Bellman, inventor of dynamic programming, had to hide the fact he was inventing it from the Secretary of Defense:…
• Twitter
  RT @chrislhayes: Former acting head of the Civil Rights Division
• White Male Terrorists Are an Issue We Should Discuss | Teen Vogue
  RT @grumpygamer: "There were over 300 mass shootings in the US in 2015, and less than 1 percent of them were committed by Muslims"

Digest powered by RSS Digest

• Google launches animation tool so you can turn boring data into cool GIFs
• Trump administration plans to minimize civil rights efforts in agencies – The Washington Post
• She’s a Climate Scientist. Here’s Why She Quit Working for Trump. | Mother Jones
• Git cheat sheet zeroturnaround.com
• Nvidia Announces Partnership To Accelerate Machine Learning-Optimized Server Deployment
• UCU report: ‘academics work two days a week unpaid’ | THE News
• Jared Kushner’s charmed life is about to come to a screeching halt | Walter Shapiro | Opinion | The Guardian
• An Overview of Multi-Task Learning for Deep Learning
• These Early ’70s Ads Tried to Convince Kids the US Army Wasn’t Totally Uptight | Mother Jones
• Twitter
  RT @ryouichiii: who wants to join the fluffy moth appreciation squad
• Twitter
  RT @votolatino: " I am 15th generation #Texan" #HeretoStay #SB4 #VLGoingLocal
• Twitter
  RT @EricaJoy: Unsurprising. Having a daughter flips the "empathy for women" bit for many dudes.
• Is China Outsmarting America in A.I.? – NYTimes.com
• Twitter
  RT @jimsciutto: Remarkable: French President Macron puts Trump in same category as Putin and Erdogan
• Twitter
  RT @chick_in_kiev: Some Women Are Not Happy About A Men-Only Committee To Reshape American Healthcare
• Dissecting Trump’s Most Rabid Online Following | FiveThirtyEight
  RT @lak_gcp: How to use semantic analysis to analyze a set of documents. #DataScience #machinelearning (forget the politics)
• Dissecting Trump’s Most Rabid Online Following | FiveThirtyEight
• Google’s Dueling Neural Networks Spar to Get Smarter, No Humans Required | WIRED
• Procedural Thinking: Understanding AI and ML in the next workforce revolution
  Sherol’s blog
• Notabilia To Do Journal – Levenger

Digest powered by RSS Digest

• But You Dont Look Sick? support for those with invisible illness or chronic illness The Spoon Theory written by Christine Miserandino – But You Dont Look Sick? support for those with invisible illness or chronic illness
• [1606.03498] Improved Techniques for Training GANs

  We present a variety of new architectural features and training procedures that we apply to the generative adversarial networks (GANs) framework. We focus on two applications of GANs: semi-supervised learning, and the generation of images that humans find visually realistic. Unlike most work on generative models, our primary goal is not to train a model that assigns high likelihood to test data, nor do we require the model to be able to learn well without using any labels. Using our new techniques, we achieve state-of-the-art results in semi-supervised classification on MNIST, CIFAR-10 and SVHN. The generated images are of high quality as confirmed by a visual Turing test: our model generates MNIST samples that humans cannot distinguish from real data, and CIFAR-10 samples that yield a human error rate of 21.3%. We also present ImageNet samples with unprecedented resolution and show that our methods enable the model to learn recognizable features of ImageNet classes.

• An introduction to Generative Adversarial Networks (with code in TensorFlow) – AYLIEN

  There has been a large resurgence of interest in generative models recently (see this blog post by OpenAI for example). These are models that can learn to create data that is similar to data that we give them. The intuition behind this is that if we can get a model to write high-quality news articles for example, then it must have also learned a lot about news articles in general. Or in other words, the model should also have a good internal representation of news articles. We can then hopefully use this representation to help us with other related tasks, such as classifying news articles by topic.

  Actually training models to create data like this is not easy, but in recent years a number of methods have started to work quite well. One such promising approach is using Generative Adversarial Networks (GANs). The prominent deep learning researcher and director of AI research at Facebook, Yann LeCun, recently cited GANs as being one of the most important new developments in deep learning:

  “There are many interesting recent development in deep learning…The most important one, in my opinion, is adversarial training (also called GAN for Generative Adversarial Networks). This, and the variations that are now being proposed is the most interesting idea in the last 10 years in ML, in my opinion.” – Yann LeCun

  The rest of this post will describe the GAN formulation in a bit more detail, and provide a brief example (with code in TensorFlow) of using a GAN to solve a toy problem.

• Exponential Linear Units (ELU) for Deep Network Learning | LinkedIn

  We introduce the “exponential linear unit” (ELU) which speeds up learning in deep neural networks and leads to higher classification accuracies. Like rectified linear units (ReLUs), leaky ReLUs (LReLUs) and parameterized ReLUs (PReLUs), ELUs also avoid a vanishing gradient via the identity for positive values. However, ELUs have improved learning characteristics compared to the units with other activation functions. In contrast to ReLUs, ELUs have negative values which allows them to push mean unit activations closer to zero. Zero means speed up learning because they bring the gradient closer to the unit natural gradient.

• Generative Adversarial Networks (GANs) in 50 lines of code (PyTorch)
• Waves Rippled Through Greenland’s Ice. That’s Ominous | Climate Central
• Coral bleaching on Great Barrier Reef worse than expected, surveys show | Environment | The Guardian
• In ‘Enormous Success,’ Scientists Tie 52 Genes to Human Intelligence – NYTimes.com
• Microsoft’s weapon in high-stakes cloud-computing battle with Amazon? Freebies | The Seattle Times
• Why the President Will Not Survive the Trump Russia Scandal
• Written/Unwritten | Patricia A. Matthew | University of North Carolina Press

  Diversity and the Hidden Truths of Tenure

• In Memoriam: Jean E. Sammet 1928-2017 | News | Communications of the ACM
• Twitter
  RT @LawrenceBlock: That’s actually the origin of the word–a spoonerism that supplanted its antecedent.
• Twitter
  RT @ClimateCentral: Waves have been detected rippling through one of Greenland’s glaciers. That’s not good.
• Waves Rippled Through Greenland’s Ice. That’s Ominous | Climate Central
  RT @ClimateCentral: Waves have been detected rippling through one of Greenland’s glaciers. That’s not good.

Digest powered by RSS Digest

• Conditional probability explained visually
• Mitch McConnell Implicated In Widespread Putin/Ukraine Money Funnel; FBI Notified
• Texas Bill Could Send People to Jail for Driving a Woman to an Abortion, Lawmaker Says
• It was quite a week for toxic masculinity | Jessica Valenti | World news | The Guardian
• Does ISIS Hate Little Girls? – Feministani by Bina Shah
• Celebrate World Gin Day With These 9 Gins… Just Not All at Once | HuffPost
• Denis Johnson in The New Yorker
• Trump White House Blinks In Standoff With Ethics Agency Over Info About Waivers For Ex-Lobbyists : NPR
• Trump’s Hand-to-Hand Combat – NYTimes.com
• Monument to peer review unveiled in Moscow : Nature News & Comment
• Brood Awakening: 17-Year Cicadas Emerge 4 Years Early – Scientific American
• Twitter
  RT @RVAwonk: Once again, I ask you to pay attention to this. At least three Americans were murdered by white supremacists THIS W…
• Twitter
  RT @bourgwick: these are amazing. never seen any pictures of pre-folkie jerry garcia playing electric guitar. #deadfreaksunite
• Twitter
  RT @DickKingSmith: Look at this shameless cat, caught in flagrante delicto with some voyeuristic hens.
• AlphaGo’s next move | DeepMind
• After 54 Years, We Fell in Love. After Five Months, I Got Leukemia. – NYTimes.com
• Trump: The Presidency in Peril | by Elizabeth Drew | The New York Review of Books
• Airbnb is running its own internal university to teach data science | TechCrunch
• Famed Hacker Kevin Mitnick Shows You How to Go Invisible Online | WIRED
• Danny Baker on Twitter: "A typical, masterly @MarinaHyde column that also manages to incorporate my favourite paragraph in all of Wodehouse. https://t.co/Zv7gjHtj6T"
• Twitter
  RT @Waterstones: ‘Fiction takes its time’: Arundhati Roy on why it took 20 years to write her second novel
• ‘Fiction takes its time’: Arundhati Roy on why it took 20 years to write her second novel | Books | The Guardian

Digest powered by RSS Digest

• Understanding deep learning requires re-thinking generalization | the morning paper
• Questions & Intuition for Tackling Deep Learning Problems
  Never mind a neural network; can a human with no prior knowledge, educated on nothing but a diet of your training data set, solve the problem? Is your network looking at your data through the right lens? Is your network learning the quirks in your training data set, or is it learning to solve the problem at hand? Does your network have siblings that can give it a leg-up (through pre-trained weights)? Is your network incapable or just lazy? If it’s the latter, how do you force it to learn?
• the morning paper | an interesting/influential/important paper from the world of CS every weekday morning, as selected by Adrian Colyer
  an interesting/influential/important paper from the world of CS every weekday morning, as selected by Adrian Colyer
• Usage patterns and the economics of the public cloud | the morning paper
• [1703.07326] One-Shot Imitation Learning
  Imitation learning has been commonly applied to solve different tasks in isolation. This usually requires either careful feature engineering, or a significant number of samples. This is far from what we desire: ideally, robots should be able to learn from very few demonstrations of any given task, and instantly generalize to new situations of the same task, without requiring task-specific engineering. In this paper, we propose a meta-learning framework for achieving such capability, which we call one-shot imitation learning.
  Specifically, we consider the setting where there is a very large set of tasks, and each task has many instantiations. For example, a task could be to stack all blocks on a table into a single tower, another task could be to place all blocks on a table into two-block towers, etc. In each case, different instances of the task would consist of different sets of blocks with different initial states. At training time, our algorithm is presented with pairs of demonstrations for a subset of all tasks. A neural net is trained that takes as input one demonstration and the current state (which initially is the initial state of the other demonstration of the pair), and outputs an action with the goal that the resulting sequence of states and actions matches as closely as possible with the second demonstration. At test time, a demonstration of a single instance of a new task is presented, and the neural net is expected to perform well on new instances of this new task. The use of soft attention allows the model to generalize to conditions and tasks unseen in the training data. We anticipate that by training this model on a much greater variety of tasks and settings, we will obtain a general system that can turn any demonstrations into robust policies that can accomplish an overwhelming variety of tasks.
• Stephen Wolfram: A New Kind of Science | Online—Table of Contents
• AgileBits Blog | Introducing Travel Mode: Protect your data when crossing borders
  1password
• Measures for Justice
  for WA
• Measures for Justice
  collects data on justice systems in several states, funded by Gates and Zuckerberg foundations. As Bach says, “Justice in America happens in 3,000 counties, each with its own justice system.”
• GitHub – oarriaga/face_classification: Real-time face detection and emotion/gender classification using fer2013/imdb datasets with a keras CNN model and openCV.
  Real-time face detection and emotion/gender classification using fer2013/IMDB datasets with a keras CNN model and openCV.

  IMDB gender classification test accuracy: 96%.
  fer2013 emotion classification test accuracy: 66%.

• We’re About to Cripple the Genomic Medical Era – NewCo Shift
  Trumpcare needlessly cedes US leadership in data, science, and health
• The Calculus of Service Availability – ACM Queue
  As detailed in Site Reliability Engineering: How Google Runs Production Systems1 (hereafter referred to as the SRE book), Google products and services seek high-velocity feature development while maintaining aggressive SLOs (service-level objectives) for availability and responsiveness. An SLO says that the service should almost always be up, and the service should almost always be fast; SLOs also provide precise numbers to define what "almost always" means for a particular service. SLOs are based on the following observation:

  The vast majority of software services and systems should aim for almost-perfect reliability rather than perfect reliability—that is, 99.999 or 99.99 percent rather than 100 percent—because users cannot tell the difference between a service being 100 percent available and less than "perfectly" available. There are many other systems in the path between user and service (laptop, home WiFi, ISP, the power grid, …), and those systems collectively are far less than 100 percent available. Thus, the marginal difference between 99.99 percent and 100 percent gets lost in the noise of other unavailability, and the user receives no benefit from the enormous effort required to add that last fractional percent of availability. Notable exceptions to this rule include antilock brake control systems and pacemakers!
  For a detailed discussion of how SLOs relate to SLIs (service-level indicators) and SLAs (service-level agreements), see the "Service Level Objectives" chapter in the SRE book. That chapter also details how to choose metrics that are meaningful for a particular service or system, which in turn drives the choice of an appropriate SLO for that service.

  This article expands upon the topic of SLOs to focus on service dependencies. Specifically, we look at how the availability of critical dependencies informs the availability of a service, and how to design in order to mitigate and minimize critical dependencies.

• GitHub – openai/baselines: OpenAI Baselines: high-quality implementations of reinforcement learning algorithms
  We’re releasing OpenAI Baselines, a set of high-quality implementations of reinforcement learning algorithms. To start, we’re making available an open source version of Deep Q-Learning and three of its variants.

  These algorithms will make it easier for the research community to replicate, refine, and identify new ideas, and will create good baselines to build research on top of. Our DQN implementation and its variants are roughly on par with the scores in published papers. We expect they will be used as a base around which new ideas can be added, and as a tool for comparing a new approach against existing ones.

• How Your Data is Stored, or, The Laws of the Imaginary Greeks
  eventual consistency VERY LUCIDLY explained. It follows the original (entertaining) paper by Leslie Lamport but spells everything out clearly for non-computer-scientists.
• [1705.07962] pix2code: Generating Code from a Graphical User Interface Screenshot
  Transforming a graphical user interface screenshot created by a designer into computer code is a typical task conducted by a developer in order to build customized software, websites and mobile applications. In this paper, we show that Deep Learning techniques can be leveraged to automatically generate code given a graphical user interface screenshot as input. Our model is able to generate code targeting three different platforms (i.e. iOS, Android and web-based technologies) from a single input image with over 77% of accuracy.
  Comments:Submitted to 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA
  Subjects:Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Neural and Evolutionary Computing (cs.NE)
  MSC classes:68T45
  ACM classes:I.2.1; I.2.10; I.2.2; I.2.6
  Cite as:arXiv:1705.07962 [cs.LG]
  (or arXiv:1705.07962v1 [cs.LG] for this version)
• GitHub – tonybeltramelli/pix2code: pix2code: Generating Code from a Graphical User Interface Screenshot
  pix2code: Generating Code from a Graphical User Interface Screenshot
• This app uses artificial intelligence to turn design mockups into source code
  RT @NandoDF: This app uses artificial intelligence to turn design mockups into source code. Cool Pixels2Code
• Me and my penis: 100 men reveal all | Life and style | The Guardian
• Interview with crime fiction author Christopher Brookmyre as he publishes his 18th novel, Black Widow (From HeraldScotland)
• Scio: Moving Big Data to Google Cloud, a Spotify Story
• Google has reportedly launched a new AI-focused venture capital program | TechCrunch
• Ancestry.com takes DNA ownership rights from customers and their relatives
• Twitter
  RT @teioh: God help us all
• Witnesses: Man Cut the Throats of Two MAX Passengers Who Tried to Stop Anti-Muslim Bullying of Women on Northeast Portland Train – Willamette Week
• Hillary Clinton Roasts the Trump Administration in a Remarkable Speech | The Nation
• Google’s AlphaGo Trounces Humans—But It Also Gives Them a Boost | WIRED
• Juno spacecraft reveals a more complex Jupiter | Science News
• Inside Hillary Clinton’s Surreal Post-Election Life
• Twitter
  RT @surt_lab: So about that "you only get your mitochondria from your mom" thing…. The answer is more complicated

Digest powered by RSS Digest

• AT&T completes its nationwide LTE-M network ahead of schedule

  Meet your new wireless network for IoT http://bit.ly/2s1KWEn

  #Tech #News #smartdevices https://twitter.com/RWW/status/868154288786337792/photo/1

• dmlc/xgboost: Scalable, Portable and Distributed Gradient Boosting (GBDT, GBRT or GBM) Library, for Python, R, Java, Scala, C++ and more. Runs on single machine, Hadoop, Spark, Flink and DataFlow
  Scalable, Portable and Distributed Gradient Boosting (GBDT, GBRT or GBM) Library, for Python, R, Java, Scala, C++ and more. Runs on single machine, Hadoop, Spark, Flink and DataFlow
• Did the Turkish President’s Security Detail Attack Protesters in Washington? What the Video Shows – The New York Times
• Calling Bullshit — Syllabus
• The GOP inherits what Trump has wrought – The Washington Post
• Mossberg: The Disappearing Computer – Recode
• Why the Atom text editor is a solid tool for writers | Opensource.com
• AlphaGo offers a sobering look into the future of man versus machine | WIRED UK
• Capturing the Earth as art | Cosmos
• Twitter
  RT @JasmineLeiylani: Hold on a second. It’s 2017 & she’s only the FIRST ASIAN WOMAN ON THE FRONT PAGE COVER OF @VanityFair WTF Vanity Fa…
• How to build a static website with Jekyll | Opensource.com
• Eating Beans Could Be A Magical Solution To Climate Change
• Barack Obama on food and climate change: ‘We can still act and it won’t be too late’ | Global development | The Guardian
• At His Own Wake, Celebrating Life and the Gift of Death – NYTimes.com
• NASA Spacecraft Finds a Chaotic Dance of Storms at Jupiter’s Poles – NYTimes.com
• Twitter
  RT @marlaerwin: I miss this.
• Twitter
  RT @NASA: Jupiter’s poles are covered in cyclones, some as big as the Earth – That & more new results from @NASAJuno. Details…
• Calling Bullshit.

  After months of work, we’re now live.

  A course for these troubled times: Calling Bullshit in the Age of Big Data.

  http://callingbullshit.org https://twitter.com/CT_Bergstrom/status/819430803385876480/photo/1

• Paleontologists with the UW’s Burke Museum discover major T. rex fossil | UW Today
• Twitter
  RT @roebuk: Client-side security at it’s best
• Uber, Lyft returning to Austin on Monday | The Texas Tribune
• Twitter
  RT @EdJoyce: 23 hours ago from @mkvackay —
  Eleven Mile state park last night! @Colorado #stormhour #colorado #Astrophotography…
• AlphaSOC on Twitter: "Our lightweight utility to submit data to the DNS Analytics API and retrieve alerts is now available. Happy hunting! https://t.co/FwLf50n4cM https://t.co/t7f6EBxRyv"
• Twitter
  RT @landpsychology: The cat No I haven’t seen him
• Twitter
  RT @karaswisher: Silicon Valley technical women’s group cuts ties with Uber, citing "continuing allegations’
• A Silicon Valley technical women’s group has cut ties with Uber, citing ‘continuing allegations’ about the treatment of female employees – Recode
• Twitter
  RT @MelissaSulewski: It reads like an Onion piece.
• Innocence Project victory: Shaurn Thomas walks free after 24 years
• Gregory Hines & Sammy Davis Jr – YouTube

Digest powered by RSS Digest

• Starting on Capitol Hill, The Riveter creating coworking spaces focused on women | CHS Capitol Hill Seattle
• Cloud+Data NextCon
  CLOUD+DATA NEXT CONFERENCE
  Innovation powered by cloud and data technology
  Silicon Valley | Jul 15-16, 2017
• A warning on Trump’s budget
  RT @FT: Larry Summers: Donald Trump’s budget includes an error that would justify failing an introductory economics course
• The Eliza Effect – Medium
  Sherol Chen
• Twitter
  RT @TopherSpiro: Remember, this is the guy who joked around looking in his wallet for the money for high-risk pools.
• Maggie Haberman Profile – How Maggie Haberman Covers Donald Trump
• Twitter
  RT @TurtlesRCool2: When your advisor asks how your research is going.
• Rage Against the Dying of the Reefs – Progressive.org
• London Review of Books
• Trump Organization Won’t Track All Foreign Government Payments – The Atlantic
• Larry Summers: Trump’s budget is simply ludicrous – The Washington Post
• Drinking coffee may help prevent liver cancer, study suggests | Science | The Guardian
• Uni-ball Signo: A Comprehensive Guide – JetPens.com
• Lamy Petrol Ink – 50 ml Bottle – JetPens.com
• Uni-ball Signo Needle UM-151ND Gel Pens – JetPens.com
• Data For Everyone Library | CrowdFlower
• Complete searchable list of Netflix Genres with links
• Twitter
  RT @AriBerman: Sessions should’ve resigned 2 months ago for lying under oath to Congress. This should be last straw
• Netflix movie codes: The secret numbers that unlock 1000s of hidden films and TV shows
• Twitter
  RT @Tobymalcolm: This little scene cheered me up today.
• Twitter
  RT @EricHolthaus: A reminder: We are quickly entering an era unseen on Earth in millions of years.
  This is the most urgent problem ou…
• Star Wars Episode IX Would’ve Been Carrie Fisher’s Movie
• Playing Maura on ‘Transparent’ Makes Jeffrey Tambor ‘Cranky’
• Nudity on Set: What It’s Like When You Remove the Male Gaze
• Twitter
  RT @SenKamalaHarris: He’s lied under oath. He’s misled on security clearance forms. It’s simple – he should not be the Attorney General.
• Why Saudi Women Are Literally Living ‘The Handmaid’s Tale’ – The New York Times
• Twitter
  RT @CassLowe: Police catch a pigeon with 200 ecstasy pills hidden in a little backpack
• For the Illiterate Adult, Learning to Read Produces Enormous Brain Changes – Scientific American Blog Network
• Twitter
  RT @ollyog: Went to a museum armed with Face App to brighten up a lot of the sombre looks on the paintings and sculptures. The…

Digest powered by RSS Digest

• Google Cloud Platform Blog: Istio: a modern approach to developing and managing microservices
  Today Google, IBM and Lyft announced the alpha release of Istio: a new open-source project that provides a uniform way to help connect, secure, manage and monitor microservices.
  Istio encapsulates many of the best practices Google has been using to run massive-scale services in production for years. We’re happy to contribute this to the community as an open solution that works with Kubernetes; on-premises or in any cloud, to help solve challenges in modern application development. Istio provides developers and devops fine-grained visibility and control over traffic without requiring any changes to application code and provides CIOs and CSOs the tools needed to help enforce security and compliance requirements across the enterprise.
• google/youtube-8m: Starter code for working with the YouTube-8M dataset.
  Starter code for working with the YouTube-8M dataset. https://research.google.com/youtube8m/
• West Big Data Innovation Hub Annual All Hands Meeting | Boulder, CO | Summary | powered by RegOnline
• Next Ride
  denver airport transport
• The truth we rarely hear: From mirror to prism | The Psychologist
• The 26 major cities with the highest quality of life in the world – Business Insider Nordic
• Morphosis: Paul Klee, ‘Dancing Under the Empire of Fear’ (1938)
• How America’s Leading Science Fiction Authors Are Shaping Your Future | Arts & Culture | Smithsonian

  (2/2) Yes, @GreatDismal called utopian/dystopian SF a “pointless dichotomy” in this (great!) piece by @Eileen_gunn: http://www.smithsonianmag.com/arts-culture/how-americas-leading-science-fiction-authors-are-shaping-your-future-180951169/

• 21st Century Yokel by Tom Cox: Unbound
• Some Thoughts About The Creation Of My New Book – Tom Cox
• Pope Francis says destroying the environment is a sin | World news | The Guardian
• Fundraiser by Jeff Lew : Erase Tacoma School Lunch Debt!
• Fundraiser by Jeff Lew : Erase Renton School Lunch Debt!
• Seattle parent pays off $21K school-lunch debt with GoFundMe campaign | The Seattle Times
• Review: In a New Book, Pope Francis Calls Mercy Essential – The New York Times
• Manchester’s heartbreak: ‘I never grasped what big pop gigs were for until I saw one through my daughter’s eyes’ | UK news | The Guardian
• Roger Moore wasn’t a good Bond, but he was my Bond – Salon.com
• The Amazing Life and Work of Maria Sibylla Merian – Atlas Obscura
• Combine K8S + Let’s Encrypt, Kubernetes Persistent Storage –
  Seattle Kubernetes Meetup (Seattle, WA)

  | Meetup

  RT @kubernetesSEA: Come & learn about using Let’s Encrypt w kubernetes and where we’re at with persistent storage next week
• Twitter
  Trump plans a 69 percent budget cut, large staff reductions at clean energy office
• Trump is targeting the government’s top renewable energy office for a 69 percent budget cut – The Washington Post
• Trump Budget Would Increase Homelessness and Hardship in Every State, End Federal Role in Community Development | Center on Budget and Policy Priorities
• Day One – Scytale.io – Medium
• SPIFFE / About
• How to Stay Sane if Trump is Driving You Insane: Advice From a Therapist
• Think Your Credentials Are Ignored Because You’re A Woman? It Could Be : 13.7: Cosmos And Culture : NPR
• Twitter
  RT @gregsramblings: @SFist Salesforce Tower from above (from my flight this morning)
• Do Donald Trump’s Old Tweets Predict His Future? – The Atlantic
• Trump Budget Based on $2 Trillion Math Error [Updated]

Digest powered by RSS Digest

• 70+ Resources for Transitioning to a Data Science Career
• Twitter
  RT @AltStateDpt: If we could retweet this thousands of times, we would.

  This is a fundamental question in a potential turning point…

• 3quarksdaily: The primacy of doubt in an age of illusory certainty
• Obama Advisor Valerie Jarrett’s Speech at Spelman College | Time.com
• Kate Heddleston on Twitter: "Omg yes!!!!! This explains so succinctly the uneven split of domestic work between men and women. https://t.co/sypvnVw5m9"
• You should’ve asked | Emma
• Netflix is creating a movie starring Rihanna and Lupita Nyong’o, thanks mostly to Twitter | TechCrunch
• Graeme Wood’s Response to Alex Tizon’s "My Family’s Slave" – The Atlantic
• shauna on Twitter: "reading my news feeds like https://t.co/kgX58hDFFM"
• Twitter
  RT @stuartpstevens: And not one car accident caused by a female driver.
• Every Color Of Cardigan Mister Rogers Wore From 1979–2001
• Twitter
  RT @debcha: The answers to this are kind of amazing. /via @quince
• Trump asked intelligence chiefs to push back against FBI collusion probe after Comey revealed its existence – The Washington Post
  RT @carlzimmer: Trump asked intelligence chiefs to push back against FBI collusion probe (They said no)
• Report reveals Uber charges riders more in wealthier areas without telling drivers | abc7news.com
• Twitter
  RT @AoDespair: When it is necessary to assure people that a man murdered at random by a stranger was "not a thug" a nation’s racia…
• Revealed: Google tried to block media coverage of gender discrimination case | Technology | The Guardian
• Sally Albright on Twitter: "“The Gaslighting of the Democrats” — Val Perry Rendel https://t.co/uJHamsrIOO https://t.co/XzeAFHKOly"
• The Rise of the Data Engineer – freeCodeCamp
• Twitter
  RT @vambenepe: Good illustration of what’s different with public datasets on Google Cloud:
  - prepared for analysis
  - ready to quer…
• Why It’s So Hard to Admit You’re Wrong – NYTimes.com
• How Zendesk Serves TensorFlow Models in Production – Zendesk Engineering – Medium
  RT @ryanseddon: How Zendesk Serves Tensorflow Models in Production –
• Panama: The Hidden Trillions | by Alan Rusbridger | The New York Review of Books
• Josh Gordon on Twitter: "Just tried YOLO for object detection in images, seems to work pretty well! Code & pretrained models: https://t.co/4whvvSgoNN https://t.co/E7rae3uNJ2"
• GitHub – thtrieu/darkflow: Translate darknet to tensorflow. Load trained weights, retrain/fine-tune using tensorflow, export constant graph def to mobile devices

  Translate darknet to tensorflow. Load trained weights, retrain/fine-tune using tensorflow, export constant graph def to mobile devices

• Twitter
  RT @francesc: "The State of Go" May 2017
  My #gopherfest slides are online #golang
• The State of Go
  RT @francesc: "The State of Go" May 2017
  My #gopherfest slides are online #golang
• Sean Hannity just gave the game away – The Washington Post
• How Women Mentors Make a Difference in Engineering – The Atlantic

  Very cool randomized, year-long study showing benefit of women mentors in engineering. https://www.theatlantic.com/science/archive/2017/05/women-mentors-engineering/527625/

• White House document promotes "lasting peach" in Middle East.

Digest powered by RSS Digest

• ML Toolkit (TensorFlow Dev Summit 2017) – YouTube
  TensorFlow is an extremely powerful framework, yet has been missing packaged solutions that work out-of-the-box. In this talk, Ashish Agarwal introduces a toolkit of algorithms that takes a step in that direction.
• broadinstitute/scalable_analytics: Public collaboration of Scalable Single Cell Analytics
  Public collaboration of Scalable Single Cell Analytics
• Twitter
  RT @mattyglesias: This only works because of Germany’s expansive, sunny deserts and wind-swept plains; would never work in the USA.
• Campaign group to challenge UK over surrender of passwords at border control | Politics | The Guardian
  RT @Pinboard: We need a travel mode for social media accounts. This man was arrested at Heathrow for not providing passwords:
• Swift is like Kotlin
• Stevey’s Blog Rants: Why Kotlin Is Better Than Whatever Dumb Language You’re Using

  .@steve_yegge on Kotlin, and why Android is its killer app: http://bit.ly/2rsUOdK

• Stevey’s Blog Rants

  .@steve_yegge on Kotlin, and why Android is its killer app: http://bit.ly/2rsUOdK

• Surfer John John Florence’s Very Wavy World | GQ
• Israeli Intelligence Furious Over Trump’s Loose Lips | Foreign Policy
• Twitter
  "White House Moves to Block Ethics Inquiry Into Ex-Lobbyists on Payroll"
• White House Moves to Block Ethics Inquiry Into Ex-Lobbyists on Payroll – The New York Times
• Golang Iota for Enumerations – ozmox
• ‘The Internet Is Broken’: @ev Is Trying to Salvage It – NYTimes.com
• 85% of Germany’s power just came from renewable energy, setting a new record | indy100

  85% of Germany’s power just came from renewable energy, setting a new record http://sc.org/2qB7fRT

• Untitled (http://www.sfgate.com/news/texas/article/Texas-Senate-approves-religious-refusal-11163562.php)
  RT @jilevin: Texas Senate approves ‘religious refusal’ adoption measure
• Texas Senate approves ‘religious refusal’ adoption measure – SFGate
• Google Home: An Insight Into a 4-Year-Old’s Mind – Jono Bacon
• Pinboard on Twitter: "Remember that Lieberman (who may become head of the FBI) wrote a bill that would give the President emergency powers over the Internet"
• Wireless Time for Jewelers (1912)
• ACLU Seeks Documents on ICE’s Use of Cell Phone Trackers | American Civil Liberties Union
• Want To Hold An ICO? CoinList Makes It Easy — And Legal
• developer-roadmap/README.md at master · kamranahmedse/developer-roadmap · GitHub
• GitHub – kamranahmedse/developer-roadmap: Roadmap to becoming a web developer in 2017
• The Skywhale – Wikipedia
• Kelsey Hightower – Keynote – Pycon 2017 – YouTube
• How To Public Speaking – YouTube
• Microsoft, Google Gaining on AWS, But Not Quickly Enough — Redmond Channel Partner
• [1611.01604] Dynamic Coattention Networks For Question Answering
  Several deep learning models have been proposed for question answering. However, due to their single-pass nature, they have no way to recover from local maxima corresponding to incorrect answers. To address this problem, we introduce the Dynamic Coattention Network (DCN) for question answering. The DCN first fuses co-dependent representations of the question and the document in order to focus on relevant parts of both. Then a dynamic pointing decoder iterates over potential answer spans. This iterative procedure enables the model to recover from initial local maxima corresponding to incorrect answers. On the Stanford question answering dataset, a single DCN model improves the previous state of the art from 71.0% F1 to 75.9%, while a DCN ensemble obtains 80.4% F1.
• CS224n: Natural Language Processing with Deep Learning
  CS224n: Natural Language Processing with Deep Learning

Digest powered by RSS Digest