Amazon’s general counsel, David Zapolsky, had a lot on his mind last month when he and four members of his legal team visited the offices of New America, a liberal-leaning think tank in Washington. The retail juggernaut was days from announcing its $13.8 billion purchase of Whole Foods, a deal that would not only roil the grocery industry but also trigger a government antitrust investigation into the strategies and practices of the “Everything Store.” And, as Zapolsky was no doubt aware, no organization had been more dogged in raising those concerns than New America — and, in particular, a 28-year-old law student named Lina Khan.
Earlier this year, the Yale Law Journal published a 24,000-word “note” by Khan titled “Amazon’s Antitrust Paradox.” The article laid out with remarkable clarity and sophistication why American antitrust law has evolved to the point that it is no longer equipped to deal with tech giants such as Amazon.com, which has made itself as essential to commerce in the 21st century as the railroads, telephone systems and computer hardware makers were in the 20th.
It’s not just Amazon, however, that animates concerns about competition and market power, and Khan is not the only one who is worrying.
As is usually the case with fast-advancing technologies, AI has inspired massive FOMO , FUD and feuds. Some of it is deserved, some of it not — but the industry is paying attention. From stealth hardware startups to fintech giants to public institutions, teams are feverishly working on their AI strategy. It all comes down to one crucial, high-stakes question: ‘How do we use AI and machine learning to get better at what we do?’
More often than not, companies are not ready for AI. Maybe they hired their first data scientist to less-than-stellar outcomes, or maybe data literacy is not central to their culture. But the most common scenario is that they have not yet built the infrastructure to implement (and reap the benefits of) the most basic data science algorithms and operations, much less machine learning.
As a data science/AI advisor, I had to deliver this message countless times, especially over the past two years.
arXiv, Computer Science > Graphics What You Sketch Is; Johanna Delanoy, Adrien Bousseau, Mathieu Aubry, Phillip Isola, Alexei A. Efros
from
Sketch-based modeling strives to bring the ease and immediacy of drawing to the 3D world. However, while drawings are easy for humans to create, they are very challenging for computers to interpret due to their sparsity and ambiguity. We propose a data-driven approach that tackles this challenge by learning to reconstruct 3D shapes from one or more drawings. At the core of our approach is a deep convolutional neural network (CNN) that predicts occupancy of a voxel grid from a line drawing. This CNN provides us with an initial 3D reconstruction as soon as the user completes a single drawing of the desired shape. We complement this single-view network with an updater CNN that refines an existing prediction given a new drawing of the shape created from a novel viewpoint. A key advantage of our approach is that we can apply the updater iteratively to fuse information from an arbitrary number of viewpoints, without requiring explicit stroke correspondences between the drawings. We train both CNNs by rendering synthetic contour drawings from hand-modeled shape collections as well as from procedurally-generated abstract shapes. Finally, we integrate our CNNs in a minimal modeling interface that allows users to seamlessly draw an object, rotate it to see its 3D reconstruction, and refine it by re-drawing from another vantage point using the 3D reconstruction as guidance. The main strengths of our approach are its robustness to freehand bitmap drawings, its ability to adapt to different object categories, and the continuum it offers between single-view and multi-view sketch-based modeling.
The Trump administration is preparing to redirect resources of the Justice Department’s civil rights division toward investigating and suing universities over affirmative action admissions policies deemed to discriminate against white applicants, according to a document obtained by The New York Times.
The document, an internal announcement to the civil rights division, seeks current lawyers interested in working for a new project on “investigations and possible litigation related to intentional race-based discrimination in college and university admissions.”
We would like to present you another wonderful student project from this summer semester created within the class Thematic Cartography, a third semester bachelor class for Spatial Planning students. The map is called Khalid’s Way and was created by Jakob Listabarth. He describes his work in the following way:
This map tells the story of Khalid’s journey from Yemen to Austria. It shows merely one personal refugee-story out of the countless other ones, which we normally only hear about through statistics, in newspapers, or from television.
Harvard has launched a major new project called “Defending Digital Democracy” designed to overcome election hacking and protect the U.S. democratic process. Is it enough to make up for Executive Office skepticism and inaction?
Historically, insurance companies see data through the prism of better analytics to make more informed underwriting or actuarial decision-making. For example, 70% of insurers believe that FinTech’s primary impact around data will be to generate deep risk insights. The same thinking also occurred in the lending industry over the last four years, but this story did not play out so well … non-traditional lending risk assessments among alternative lenders has been the reason a number of these firms (e.g. CAN Capital) have got themselves into trouble.
One area that has significant potential is to use data to help with the customer acquisition process. Traditionally, insurance agents have relied on relationship selling supported by lead generation tools (e.g. Dunn & Bradstreet). Now, new tools exist to help insurance carriers start to predict customer needs for insurance products.
Are you interested to see how the top placing teams came about their solutions to claim their top ranking place in the 2017 Data Science Bowl? Check out their code via the links below.
Heating, venting and air-conditioning may not be a sexy application of artificial intelligence, but there is huge cost-saving potential in using AI in HVAC
Artificial intelligence (AI) seems to be everywhere. It’s been tested and has proven efficient by using AI to play backgammon, chess, the game of Go and even Atari games. In some ways, AI is catching up and overtaking us. One AIapplication that may not sound sexy, but where there is huge potential to apply AI is the field of HVAC – that’s right heating, venting and air-conditioning. HVAC systems are underappreciated technologies. They fall into the category of technology that a person uses every day and would hate to live without.
So, is there a way to apply the same smart control algorithms that have proven efficient in playing games to a commercial HVAC system that requires coordination of hundreds of control loops? To answer this, let’s start with simulation.
The term “birdbrain” has been part of our lexicon for about a century to describe someone’s intellect, or lack of it, presumably because birds have really small brains. But, in fact, the way songbirds learn to sing is similar to how humans learn speech.
A unique interdisciplinary team of Florida State University researchers is leading the way among scientists worldwide to understand the question of how the male zebra finch learns its songs. Now, with the help of an $800,000 grant from the National Science Foundation, they are moving ahead with research to answer that question.
“Birdsong is a very good model for us to understand how the human brain works in terms of vocalization and how we learn speech,” said Wei Wu, associate professor of statistics. “It’s a perfect model.”
Research and new technologies have dramatically expanded the amount of data that can be captured from the brain over the past few years. The challenge now is to determine what that data can tell us.
A group of researchers, led by Krešimir Josić, a mathematical biologist at the University of Houston, has been awarded a $4.39 million grant from the National Science Foundation to develop new methods to analyze and interpret neural data. The five-year grant was issued through the NSF NeuroNex program, or Next Generation Networks for Neuroscience.
“The amount of available data has expanded enormously,” said Josić, a professor of mathematics with joint appointments to the UH Department of Biology and Biochemistry and at Rice University. “Now that we have it, the question is, what do we do with it? The important thing isn’t the data. It is understanding what the data mean to us.”
EurekAlert! Science News, American Statistical Association
from
Although most people don’t broadcast in advance their intention to engage in criminal activity, University of Virginia Assistant Professor of Systems and Engineering Information Matthew Gerber has discovered that the use of Twitter can help predict crime. Gerber’s research and work developing statistical crime prediction methods will be presented on Tuesday, August 1, 2017, at the Joint Statistical Meetings in Baltimore, Md.
“My initial hypothesis was that there would be no correlation between Twitter use and crime. After all, people don’t share with the world that they intend to or have just committed a crime,” said Gerber. “What they do share are things like social events or outings that could lead to criminal activity.” Gerber chose Twitter over other social media platforms for its openness and the fact that anyone can access GPS-tagged tweets generated in a given area.
In this boot camp, students learn to take risks and get “messy” — by studying grammar, that is.
Claire Bowern, professor of linguistics, conceived of the Grammar Boot Camp several ago as a way of contributing new linguistic knowledge about endangered languages. Students in the program come to Yale for a month to work with archival field notes and recordings in an effort to create a publishable sketch of a grammar. This year, the program included students from the University of California-Berkeley, Carleton College, and McGill University. They were tasked with studying Noongar, an Australian Aboriginal language.
“The whole point of the boot camp is to analyze things that haven’t been analyzed before,” says Bowern.
Research Data Alliance is now seeking candidates for its annual TAB election, to be held in September 2017. Terms are for three (3) years. We encourage participation from people in all stages of their careers. Nominations will be accepted until August 15, 2017.
On September 1 Kaggle will announce the winners of the Datasets Award [and give the winners $10,000]. Any datasets published between August 1-31 will be considered for the award.
Both the Snap Research Scholarship and Fellowship consist of a $10,000 award prize and the opportunity to intern at Snap Research for the 2018 summer or fall. Deadline to apply is October 31.