Scientists at Facebook’s A.I. research lab actually think they can improve A.I.’s mastery of communication by having it guide a virtual tourist through New York City. They’ve dubbed this project Talk the Walk and they tell Inverse this method could be the key to teaching machines how to master the nuance of language.
“The goal of the research project is to explore a more human-like approach to language learning,” explains University of Montreal Ph.D. student, Harm de Vries. “Previous work mostly focused on language learning from large [collections of text], like Wikipedia. [This] project aims to explore language learning in a contextual environment, making it possible to link words and concepts to perception and action in that environment.”
Ride-sharing operator Grab will tap data from its platform to extract insights on transport patterns in Southeast Asian cities, and jointly develop applications with National University of Singapore to “transform” urban transportation.
Computer scientists must identify sources of bias, de-bias training data and develop artificial-intelligence algorithms that are robust to skews in the data, argue James Zou and Londa Schiebinger.
Some 16 months after Ford initially announced its intention to invest $1 billion in a startup with only a handful of people on its staff, remarkably little real information has been revealed about Argo AI. Based in Pittsburgh, home of Carnegie Mellon University and some of the most advanced robotics research in the world, Argo is tasked with creating a production-ready automated driving system for Ford.
Two weeks ago, Argo moved into shiny new offices located on the bank of the Allegheny River, next door to where the company started in late 2016. The engineering team currently occupies the top floor of a five-story building. They eventually expect to fill half of the structure as the company grows. As of this month, Argo has 338 employees divided among Pittsburgh; Dearborn, Michigan; Mountain View, California; and Miami, along with an office in Cranbury, New Jersey, where the lidar specialists from Princeton Lightwave, a recent Ford acquisition, are based.
The vast majority of active MI researchers and practitioners are in North America, Europe, and Asia while large regions, including Africa, are hardly represented. We have already seen how algorithmic biases can lead to inadvertent discrimination against large sections of the world’s population. Furthermore, the lack of MI researchers from Africa means that many opportunities to use MI to create a better and more stable world are being missed. The challenges we choose to work on are strongly influenced by our backgrounds and our environment. If Africa continues to be bypassed by MI, a rare opportunity to alleviate global social and economic disparities will be missed.
Conversely, creating an effective, globally connected community of MI practitioners in Africa will reduce the technology gap, strengthen Africa’s economies and enable better governance. Fortunately, many of the ingredients are already in place: (1) Africa is home to the youngest and fastest growing population on earth. Currently, 60 percent of Africans are under 25 (compared to 32 percent in North America). This is a vast pool of talent whose enthusiasm for MI is well illustrated by the success of the Black in AI organization and summer schools such as Data Science Africa (DSA) and Deep Learning Indaba. (2) The ideal vehicle already exists. The African Institute for Mathematical Sciences (AIMS) is a network of centers of excellence spanning the African continent. AIMS has been recruiting and training Africa’s top graduates in maths, science and engineering for the past 15 years.
One of the most powerful examples of data visualization was made 118 years ago by an all-black team led by W.E.B. Du Bois only 37 years after the end of Slavery in the United States.
While Du Bois’ legacy is cemented in American history, his data visualizations remain relatively unknown. This four-part series will explore many of the individual charts from a design perspective and will cover:
I. The Exhibit of American Negroes
II. Data Journalism and the scientific study of “The Negro Problem”
III. Design, innovation and statistical chart making in 1899
IV. Du Bois as Social Scientist and the Legacy of “The Exhibit of American Negroes”
Federal Trade Commission members said during a House hearing Wednesday that the FTC does not have the authority to adequately punish companies that misuse consumer data.
Several data privacy scandals over the last year and a half prompted the House Energy and Commerce Committee hearing.
Equifax’s data breach in 2017, for example, released personal information on roughly 143 million Americans, including Social Security and driver’s license numbers. And Facebook’s Cambridge Analytica scandal in March sparked a new wave of public distrust in social media after the Republican-affiliated firm gathered user data on 50 million accounts without permission.
Google was hit with a $5.1 billion fine by the European Union on Wednesday for antitrust practices around the Android mobile operating system – a move that underscores Europe’s willingness to issue steep financial penalties for bad behavior.
And there’s another bludgeon in the EU’s cache that Google – and the advertising world in general – have to fear: the General Data Protection Regulation (GDPR).
But nearly two months past the General Data Protection Regulation compliance deadline, regulators in Europe are still playing their cards close to the vest.
The stakes could hardly be higher, as GDPR permits fines of up to 20 million euros or 4% of annual global turnover.
Jackson is currently at work on what he describes as “not the usual film that you would expect on the First World War,” and as part of that documentary work, he has digitally enhanced footage from the period, “incredible footage of which the faces of the men just jump out at you. It’s the faces, it’s the people that come to life in this film. It’s the human beings that were actually there, that were thrust into this extraordinary situation that defined their lives in many cases.” In addition to restoring old film, Jackson and his team have combed through about 600 hours of audio interviews with WWI veterans, in order to further communicate “the experience of what it was like to fight in this war” from the point of view of the people who fought it.
GE (NYSE: GE) and Microsoft Corp. (Nasdaq: “MSFT”) today announced an expanded partnership, bringing together operational technology and information technology to eliminate hurdles industrial companies face in advancing digital transformation projects. As part of the union, GE Digital plans to standardize its Predix solutions on Microsoft Azure and will deeply integrate the Predix portfolio with Azure’s native cloud capabilities, including Azure IoT and Azure Data and Analytics. The parties will also co-sell and go-to-market together, offering end customers premier Industrial IoT (IIoT) solutions across verticals. In addition, GE will leverage Microsoft Azure across its business for additional IT workloads and productivity tools, including internal Predix-based deployments, to drive innovation across the company.
If rumors are to be believed, Amazon Web Services is building its own networking gear, presumably to sell to customers that want a hardware-based foundation on which to run a hybrid cloud environment. True or not, I think it’s a smart idea (and one I’m pretty sure I’ve advocated here before) as a way to truly own the hybrid cloud story — a story that, despite years of public-cloud-only rhetoric, absolutely resonates with many enterprise CIOs.
A hardware-based approach to hybrid cloud is smart for two main reasons: (1) It gives the cloud provider the opportunity to optimize the entire package to its own specs (a la Apple products), and (2) cloud providers already know a ton about designing hardware and getting it built on the cheap. If, more than a decade into the cloud computing era, you still can’t bring every workload over, then perhaps it’s a good idea to take what you’ve learned about hardware and data centers and go to those workloads.
Artificial intelligence poses a threat to a lot of jobs out there, but Promethean AI thinks it can help game artists do their jobs better through automation of the tasks that are either too boring or repetitive.
Andrew Maximov, former technical art director at Naughty Dog, worked as a lead artist on games such as the Uncharted series, where task of building out virtual landscapes required hundreds of artists and could be overwhelming at times. So he started the Los Angeles company to use AI to assist artists in the processing of building virtual worlds.
To predict when morning traffic is likely to grind to a halt, it may be more effective to examine how we use electricity in the middle of the night instead of travel-time data.
When wives earn more than their husbands do, a puzzling thing can happen: Husbands say they earn more than they are and wives underreport their income.
New Census Bureau research shows that the incomes couples report on Census Bureau surveys do not always match their IRS filings. The Census Bureau is working to improve the quality of reported earnings by comparing an individual’s survey response with their reported response from another source.
“This year ICLR will not have a separate Workshop Track. All submissions will be to the main conference track. The conference will instead feature a small number of concurrent workshop sessions, which will be run through a separately chaired process similar to other ML venues.” Deadline for submissions is September 27.
“Using Keras to train a convolutional neural network to classify physical activity. The dataset was built from the recordings of 30 subjects performing basic activities and postural transitions while carrying a waist-mounted smartphone with embedded inertial sensors.”
“The update for the entertainment creation tool will help developers create more realistic characters, immersive environments, and use them across games, film, TV, virtual reality, mixed reality, augmented reality, and enterprise applications. The new engine combines real-time rendering advancements with improved creativity and productivity tools.”