Data Science newsletter – November 9, 2020

GROUP CURATION: N/A

The Harvard Data Science Initiative is pleased to announce the appointment of Sheila Jasanoff, Pforzheimer Professor of Science and Technology Studies, Harvard Kennedy School, as Faculty Lead of the HDSI’s Trust in Science project. Professor Jasanoff is a pioneer in the field of Science and Technology Studies and brings to her role decades of scholarship examining the role of science in society. She is the author of more than 130 articles or book chapters and is editor or author of more than 15 books on related topics.

Apple on Thursday advised developers they need to clarify the privacy practices of apps distributed through its App Store, a requirement previewed earlier this year.

The iPhone maker said software makers can now start creating standardized summaries, via its App Store Connect interface, that clearly define to users what kinds of personal data their code collects and handles. These details will be displayed on each app’s App Store product page next month.

“This information will be required to submit new apps and app updates to the App Store starting December 8, 2020,” Apple said.

EPFL engineers have developed a computer chip that combines two functions – logic operations and data storage – into a single architecture, paving the way to more efficient devices. Their technology is particularly promising for applications relying on artificial intelligence.

The considerable battery life required for these high-cost bio-logging systems has proved limiting so far. “Since bio-loggers attached to small animals have to be small and lightweight, they have short runtimes and it was therefore difficult to record interesting infrequent behaviours,” explains study corresponding author Takuya Maekawa.

By using AI-assisted bio-loggers, researchers can use low-cost sensors to automatically detect behaviours of interest in real time, allowing them to conditionally activate high-cost (i.e., resource-intensive) sensors to target those behaviours.

The shopping lists we used to scribble on the back of an envelope are increasingly already known by the supermarkets we frequent.

Firstly via the loyalty cards we scan at checkouts, and more and more so from our online baskets, our shopping habits are no longer a secret.

But now more retailers are using AI (artificial intelligence) – software systems that can learn for themselves – to try to automatically predict and encourage our very specific preferences and purchases like never before.

Retail consultant Daniel Burke, of Blick Rothenberg, calls this “the holy grail… to build up a profile of customers and suggest a product before they realise it is what they wanted”.

Silentium has introduced advanced, broad-band active road noise cancellation to the auto industry for the first time.

After several years in development, Jaguar and Land Rover are the first carmakers to integrate Silentium’s ‘Active Acoustics’ software in three of their new vehicles, meaning the technology is now available for car buyers to experience. Active road noise cancellation removes 90% of unwanted noise across a broad band of frequencies – from 20Hz up to 1kHz – providing a quieter and more refined experience for occupants, and therefore preventing driver fatigue.

Sometime toward the end of the last ice age, a gray wolf gingerly approached a human encampment. Those first tentative steps set his species on the path to a dramatic transformation: By at least 15,000 years ago, those wolves had become dogs, and neither they nor their human companions would ever be the same. But just how this relationship evolved over the ensuing millennia has been a mystery. Now, in the most comprehensive comparison yet of ancient dog and human DNA, scientists are starting to fill in some of the blanks, revealing where dogs and humans traveled together—and where they may have parted ways.

“It’s a really cool study,” says Wolfgang Haak, an archaeogeneticist at the Max Planck Institute for the Science of Human History. “We’re finally starting to see how the dog story and the human story match up.”

Casper Harteveld and Giovanni Troiano, are in the middle of an ongoing study that fuses elements of game design and storytelling to allow for a wide range of answers.

Based on the responses they’ve received since the study launched in July, Troiano describes paranoia, anxiety, and visions of dystopia as common themes. This, he says, did not surprise him or Harteveld because of the overall increase in feelings of fear, grief, and anxiety among the general populace.

The sheer volume of radicals — who could in 24 hours turn from a sedentary ISIS internet fanboy into a hire-car driver mounting the kerb with a machete — is an insurmountable manpower drain. The new head of the UK’s MI5 security service, Ken McCallum, warned last month that the adherents to ISIS ideology number in the tens of thousands, and they must find the few “who at any given moment might be mobilizing towards attacks. Having someone ‘on our radar’ is not,” he warned, “the same as having them under detailed real-time scrutiny.”

Facebook is pressing NYU researchers to stop collecting data on political ad targeting amid the presidential election, setting up a precedent to discourage independent investigations on Facebook without the company’s approval.

The controversy arose over Ad Observer, a web plugin that copies the political ads that users see on Facebook and puts them in the Ad Observatory public database with the aim of increasing transparency on ad targeting.

“We understand the intent behind your tool. However, the browser plugin scrapes information in violation of our terms, which are designed to protect people’s privacy,” reads an email from Facebook sent to lead researcher and P.h.D candidate Laura Edelson, who shared the email with NYU Local.

The Linley Fall Processor conference, which is taking place as a virtual event this week and next week, is one of the main meet-and-greet events every year for promising young chip companies.

To kick off the show, the conference host, Linley Gwennap, who has been a semiconductor analyst for two decades, offered a keynote Tuesday morning in which he said that software remains the stumbling block for all companies that want to challenge Nvidia’s lead in processing artificial intelligence.

“Although several chip vendors and cloud-service vendors have developed impressive hardware for AI acceleration, the next hurdle is the software,” said Gwennap

Legal writing has a reputation, and not for easy reading. A century ago, writer Franz Kafka imagined “the Law” as part nightmare, part cultural unconscious: irrational, enigmatic, and stubbornly invulnerable to critique. Anyone who watches the Sunday news shows or listens to AM talk radio knows that something has changed. Constitutional law in the United States today is hotly debated—a juridical wrestling match complete with opposing, media-branded schools of interpretation. Meanwhile, the budget of the United States Supreme Court alone exceeds $84 million annually. As a nation, we invest a lot of money and expertise to clear up legal ambiguities and to decide how laws—written months, decades, or centuries ago—should be applied to present-day situations.

Can we avoid all that? According to James Grimmelmann, Law, computer programmers and legal thinkers have occasionally shared a common dream: What if laws employed the hard-and-fast logic of C++? Imagine courts that run with the accuracy and efficiency of a supercomputer. One circuit court judge, commenting on the Supreme Court’s 2020 Bostock decision, voiced his ideal of jurisprudence like this: “A computer programmer may write faulty code, but the code will perform precisely as written regardless of what the programmer anticipated. Courts, no less than computers, are bound by what was typed, and also what was mistyped.”

“That’s a reasonably common view about legal language and programming that you find among some people in law,” Grimmelmann says. “There’s this idea of a computer that is rigid and literal—and perhaps the law could be more like that.”

Dr. Abhishek Dubey was announced as a recipient of grants from the NSF and the U.S. Department of Energy to research a Smart Transit system for the Chattanooga area.

Dealing with a global pandemic has taken a toll on the mental health of millions of people. A team of MIT and Harvard University researchers has shown that they can measure those effects by analyzing the language that people use to express their anxiety online.

Using machine learning to analyze the text of more than 800,000 Reddit posts, the researchers were able to identify changes in the tone and content of language that people used as the first wave of the Covid-19 pandemic progressed, from January to April of 2020. Their analysis revealed several key changes in conversations about mental health, including an overall increase in discussion about anxiety and suicide.

“We found that there were these natural clusters that emerged related to suicidality and loneliness, and the amount of posts in these clusters more than doubled during the pandemic as compared to the same months of the preceding year, which is a grave concern,” says Daniel Low, a graduate student in the Program in Speech and Hearing Bioscience and Technology at Harvard and MIT and the lead author of the study.

Oregon State University’s TRACE Community COVID-19 testing project is heading to Eugene this weekend, where collaboration with the University of Oregon will provide a better understanding of the virus’ community-wide prevalence.

Three-person TRACE Community teams will visit 30 neighborhoods in the city to collect nasal samples from as many as 600 Eugene residents.

Online November 20, starting at 12 p.m. Eastern time. “The panel will feature thought leaders from higher education and the industry, including Charles Isbell, dean of Computing at Georgia Tech and a ML@GT faculty member. Isbell will be joined by Georgia Tech alumnus Rapha Gontijo Lopes, co-founder of Queer in AI, organizer of Radical AI and AI resident at Google Brain, as well as Tiffany Deng, program management lead for machine learning fairness and responsible artificial intelligence (AI) at Google. The panel will be moderated by ML@GT Associate Director Deven Desai.” [registration required]

Online November 12, starting at 12 p.m. Eastern time. “Next up in our seminar series: our inaugural Rising Star Spotlight talks on Thurs Nov 12, featuring Irene Chen @irenetrampoline
and Arpita Biswas. One of the goals of our initiative is to highlight early career researchers; we are so excited to showcase their work!” [live streamed on YouTube]

Online November 19-20. “The Common Fund Library of Integrated Network-based Cellular Signatures (LINCS) program was launched in 2010. Its focus is perturbational biology undertaken at scale to create community resources for query and analysis. By using a multi-omic approach, LINCS investigators have built a catalog of cellular responses (signatures) from various assays (transcriptomics, proteomics, imaging) performed on a variety of different cell types following genetic, small molecule, antibody, or microenvironment perturbations. … The symposium will also include presentations from industry and academic users of LINCS data and tools.” [registration required]

What’s increment.me? Answer: An insanely simple way to build anonymous feedback into everything. Based out of Austin, Texas, Increment was created out of the belief that a growth mindset benefits the individual and the community – and we strive to help both individuals and communities thrive by creating tools that help you be your best.