Jason Millar, a postdoctoral research fellow in engineering and ethics, is one of a dozen scientists and researchers from across Stanford tackling some of the prickliest questions keeping automakers up at night.
“It is one thing to say that a machine is working better than the average human,” Millar says, “but that is based on what it means to be better. The set of criteria you use is negotiable. It can be contentious. That is where some of the interesting ethical questions arise.”
Millar wants to equip engineers to make some of the most important decisions about safety in the design room. He predicts that engineering ethics will become an important new field. “We need to get engineers solving these problems who are attuned to human ethics and human knowledge,” he says.
The fact that a mathematician could be considered, as he is, a “rock star” — or, better yet, “the Lady Gaga of mathematics” — says perhaps more about the French than [Cedric] Villani. Nonetheless, Villani, 44, has become a darling of President Emmanuel Macron’s young technocratic government, accompanying the president to Ouagadougou, Burkina Faso, in November and Beijing in mid-January. The government has piled the work on his desk, which is evidence, Villani says, of the need for people with scientific expertise in politics. But of all his projects — from math education to the future of New Caledonia to tax evasion — perhaps his most all-consuming mission is his task force on artificial intelligence and the highly anticipated report it’s set to release tomorrow. If successful, the report will help set the AI agenda in France and Europe for years to come.
In view of a world where “artificial intelligence will be everywhere, like electricity,” as Villani has said, becoming a leader in the field is critical for France.
Both car OEMs and tier ones’ appetite for software expertise — deep learning in particular — has only grown over the last 18 months. The industry, in general, suffers a chronic knowledge gap in deep learning and how to leverage it to develop software.
Even DeepScale, co-founded by Iandola, a PhD from U.C. Berkeley working on deep neural networks and computer vision systems, feels pressed to internally scale up its expertise more quickly to meet the external demands.
Asked how DeepScale intends to use the $15 million, Iandola told EE Times, “We not only need to hire deep learning experts but to develop an internal [deep learning] training program to scale the team.”
Over the past several years, Caltech mathematician Matilde Marcolli, together with her students, has begun developing new computational methods that allow her to build and analyze linguistic trees—and, specifically, to hone in on their oldest sections. To do this, Marcolli is applying several different mathematical methods to the study of language: algebraic geometry, topology, and coding techniques, among others.
“Individually, some of these methods have been applied before,” she says. “But in trying to tackle the structure of syntax in natural languages, you need a broad combination of different mathematical approaches.”
Three thousand Google employees signed a letter vowing never to work on products that are, or could be, instruments of war. They urged CEO Sundar Pichai to back out of a Pentagon partnership that could be used to “improve” drone warfare. Increasingly, companies like Google and Cambridge Analytica are realizing that intelligent, talented employees are such assets!
Twitter put out its 12th Transparency Report in which it celebrated having taken down 8% fewer terror-linked accounts than in previous time periods. They believe it’s because people who promote terror do not see Twitter as a friendly platform for terrorism. This concentrates terroristic threats on other platforms which is…still not great for the world, even if it’s good for Twitter. Abusive accounts also get taken down. This time around it was “impersonation (66%), harassment (16%), and hateful conduct (12%)” leading the top abuse take-down list.
But…there’s (another) Twitter lookalike on the scene: Mastodon is free, open-source, targets creative coders, data scientists, and visualizers, artists, curators, and critics, and gives posters 500 characters of free expression.
Slack has changed its privacy policy to allow employers the ability to read all of their employees’ DMs. Additionally, the platform used to alert users when their employers had downloaded communications logs (of any kind), but the platform isn’t going to do so anymore. Companies, please don’t go the stealth route. If you plan to read DMs, tell your employees that they shouldn’t put anything there they don’t want you to see. Remind them frequently, not just once in their employment contract. Be the difference between surveillance and stalking.
SenseTime Group Ltd. is the world’s most highly valued AI company, at $3bn. Based in China and rooted in a completely different cultural norm of surveillance, the company has apps in over 100 million phones and cameras placed all over China’s cities, capturing facial images, identifying individuals, and linking them to a variety of other types of data about these individuals. The Chinese police already use the images to find people in crowds who have warrants out for their arrests. Civil liberties groups are displeased. Co-founder Xu Li said the data, “will not affect privacy because only authorized persons can access it”. I’ll let you contemplate your own reaction to that sentiment.
Configuring high performance computing at the hardware/software interface is complicated. Most academics don’t have to figure that out on a day-to-day basis. For that reason, I’m always interested in accessible writing on the topic, and was pleased that NVidia made Ian Buck, who runs the Tesla accelerated computing business for them, available to do an interview. He has found that in HPCs, “about 70 percent of the processing cycles” are dominated by 15 applications. He spends most of his time optimizing those top 15, whatever they may be. The interview is lengthy, worth a read, and I’ll flag that he talks about “revising and adapting our architectures for the future of deep learning” as well as what they want out of their partnership with IBM and partnerships in general.
Apple is planning to make its own chips starting in 2020, ditching Intel. This will help the company build AI products that work on-device without destroying battery life.
A.I.-driven hardware produced by Microsoft is on the horizon according to Harry Shum, executive vice president of the company’s A.I. and Research Group. It’s part of Microsoft’s initiative to integrate artificial intelligence into every product and service offered by the company. Shum says these devices will be “very, very exciting.”
The comment arrives after Microsoft CEO Satya Nadella announced yet another reorganization within the company, with the result creating two new engineering teams: the Experiences & Devices group led by Rajesh Jha, and the Cloud and A.I. Platform group led by Jason Zander. Harry Shum will continue to run the current A.I. and Research group.
The conversation started over a fence dividing two backyards. On one side, an ecologist remarked that surveying animals is a pain. His neighbor, an astronomer, said he could see objects in space billions of light years away.
And so began an unusual partnership to adapt tools originally developed to detect stars in the sky to monitor animals on the ground.
The neighbors, Steven Longmore, the astronomer, and Serge Wich, the ecologist, both of Liverpool John Moores University in England, made their backyard banter a reality that may contribute to conservation and the fight against poaching.
The scientists developed a system of drones and special cameras that can record rare and endangered species on the ground, day or night. Computer-vision and machine-learning techniques that help researchers study the universe’s oldest and most distant galaxies can now be used to find animals in video footage.
Claire Burke, an astrophysicist at the university now leading the project, presented the team’s latest findings at the European Week of Astronomy and Space Science on Tuesday.
Researchers at the University of Pittsburgh and Carnegie Mellon University will join forces on a $2.25 million Defense Department study to combine human insight to the computational power of machine learning to forecast potential upheavals around the world, university officials announced Thursday.
A Pitt spokesman said researchers hope to come up with models similar to those used by meteorologists to forecast weather.
But they’ll add complex socioeconomic and geopolitical factors in an effort to predict the outcome of world events as varied as a national currency devaluation or severe weather patterns.
Researchers’ first challenge will be to predict food shortages in South Sudan, according to a release announcing the project by Natasa Miskov-Zivanov, assistant professor of electrical and computer engineering at Pitt’s Swanson School of Engineering.
The National Institutes for Health (NIH) are on an ambitious effort to harness advances in data science, machine learning, and artificial intelligence (AI) to support programs like the Precision Medicine, Cancer Moonshot, and Brain Initiatives. To accelerate progress, the NIH made a call to the public for a Request for Information (RFI)on the proposed Strategic Plan on Data Science. I submitted my letter and a number of people asked me to make my letter public. Since, as soon as it is submitted, it becomes part of the public record and the submission has now closed, I’m put the full text below.
Apple Inc. is planning to use its own chips in Mac computers beginning as early as 2020, replacing processors from Intel Corp., according to people familiar with the plans.
The initiative, code named Kalamata, is still in the early developmental stages, but comes as part of a larger strategy to make all of Apple’s devices — including Macs, iPhones, and iPads — work more similarly and seamlessly together, said the people, who asked not to be identified discussing private information. The project, which executives have approved, will likely result in a multi-step transition.
The shift would be a blow to Intel, whose partnership helped revive Apple’s Mac success and linked the chipmaker to one of the leading brands in electronics. Apple provides Intel with about 5 percent of its annual revenue, according to Bloomberg supply chain analysis.
In the two months since Richard Watson strapped 200 remote-control-sized transmitters around his cows’ necks, an artificial-intelligence system named Ida has pinged his phone with helpful alerts: when his cows are chewing the cud, when they’re feeling sick, when they’re ready for insemination.
“There may be 10 animals out there that have a real problem, but could you pick them?” he said one morning, standing among a grazing herd of dairy cattle wearing what he calls “cow Fitbits.”
But on neighboring pastures here in rural Georgia, other farmers say they aren’t that impressed. When a cow’s in heat, they know she’ll start getting mounted by her bovine sisters, so they apply a streak of paint on the cows’ backsides and then just look for the incriminating smudge. No fancy AI required.
A person watching videos that show things opening — a door, a book, curtains, a blooming flower, a yawning dog — easily understands the same type of action is depicted in each clip.
“Computer models fail miserably to identify these things. How do humans do it so effortlessly?” asks Dan Gutfreund, a principal investigator at the MIT-IBM Watson AI Laboratory and a staff member at IBM Research. “We process information as it happens in space and time. How can we teach computer models to do that?”
Such are the big questions behind one of the new projects underway at the MIT-IBM Watson AI Laboratory, a collaboration for research on the frontiers of artificial intelligence. Launched last fall, the lab connects MIT and IBM researchers together to work on AI algorithms, the application of AI to industries, the physics of AI, and ways to use AI to advance shared prosperity.
The Moments in Time dataset is one of the projects related to AI algorithms that is funded by the lab. It pairs Gutfreund with Aude Oliva, a principal research scientist at the MIT Computer Science and Artificial Intelligence Laboratory, as the project’s principal investigators. Moments in Time is built on a collection of 1 million annotated videos of dynamic events unfolding within three seconds. Gutfreund and Oliva, who is also the MIT executive director at the MIT-IBM Watson AI Lab, are using these clips to address one of the next big steps for AI: teaching machines to recognize actions.
If you want to see the future of artificial intelligence, you don’t need to head to Silicon Valley’s tech titans – just walk on to a university campus.
Many of the most exciting developments in AI are embedded in and around research-intensive universities. They are a prime source of talent, discovery and innovation. This bodes well for the future of a field that will be shaped by a vast array of startups, academic researchers and students, and not just a handful of corporate giants.
This diverse set of players, where universities play a central role, may allow AI to avoid some of the pitfalls that have afflicted internet technologies. Just weeks before Mark Zuckerberg acknowledged Facebook’s role in the Cambridge Analytica scandal, World Wide Web inventor Tim Berners-Lee bemoaned the “concentration of power [that] creates a new set of gatekeepers, allowing a handful of platforms to control which ideas and opinions are seen and shared” on the web. Universities are helping preserve this diversity in AI.
“We’re launching a transfer learning contest that measures a reinforcement learning algorithm’s ability to generalize from previous experience. In typical RL research, algorithms are tested in the same environment where they were trained, which favors algorithms which are good at memorization and have many hyperparameters. Instead, our contest tests an algorithm on previously unseen video game levels. This contest uses Gym Retro, a new platform integrating classic games into Gym, starting with 30 SEGA Genesis games.” Deadline for submissions is June 5.
“The STEM+C program supports research and development proposals related to new approaches to pre-K-12 STEM teaching and learning related to Harnessing the Data Revolution, Convergence Research and the Future of Work at the Human-Technology Frontier.” Deadline for proposals is July 2.
“We’re launching AWS Secrets Manager which makes it easy to store and retrieve your secrets via API or the AWS Command Line Interface (CLI) and rotate your credentials with built-in or custom AWS Lambda functions.”
“We’re excited to release FlameScope: a new performance visualization tool for analyzing variance, perturbations, single-threaded execution, application startup, and other time-based issues.”
“This book presents an accessible introduction to data-driven storytelling. Resulting from unique discussions between data visualization researchers and data journalists it offers an integrated definition of the topic, presents vivid examples and patterns for data storytelling, and calls out key challenges and new opportunities.”
Unique source of knowledge resulting from discussions between data visualization researchers and data journalists
State-of-the-art in data journalism and data visualization
Research agenda and opportunities for data-driven storytelling and visualization research
Curated compilation of real examples and discussion of best practices
ETL stands for Extract, Transform, Load. It has been a traditional way to manage analytics pipelines for decades. With the advent of modern cloud-based data warehouses, such as BigQuery or Redshift, the traditional concept of ETL is changing towards ELT – when you’re running transformations right in the data warehouse. Let’s see why it’s happening, what it means to have ETL vs ELT, and what we can expect in the future.
