Data Science newsletter – November 13, 2019

GROUP CURATION: N/A

… For me, narrative economics means studying the popular narratives that underlie people’s thinking, not just economists’ thinking. We have to know why people spend more in a boom and less in a bust. Why do they pay attention to the stock market? Why do they perceive the stock market as a barometer that measures the health of the economy? Why do they place such value on home ownership? Those are all parts of economic narratives. I want to study the narratives through time and across space. These ideas color people’s thinking and actions.

One of the book’s key propositions is that economic events are substantially driven by contagious spread of oversimplified and easily transmitted variants of economic narratives. We need to incorporate the contagion of these narratives into economic theory. Otherwise, we’re overlooking a very important mechanism for economic change. We need to better understand the epidemics of popular narratives in order to fully understand fluctuations in the economy and in economic behavior.

We’ve updated our analysis with data that span 1959 to 2012. Looking at the data as a whole, we clearly see two distinct eras of training AI systems in terms of compute-usage: (a) a first era, from 1959 to 2012, which is defined by results that roughly track Moore’s law, and (b) the modern era, from 2012 to now, of results using computational power that substantially outpaces macro trends. The history of investment in AI broadly is usually told as a story of booms and busts, but we don’t see that reflected in the historical trend of compute used by learning systems. It seems that AI winters and periods of excitement had a small effect on compute used to train models[2]

AiCure, a nearly decade-old startup using artificial intelligence to measure medication adherence and other behavior during clinical trials or normal care, has brought in $24.5 million in Series C funding.

Want to buy a house by the sea? If your neighbors are climate deniers, you’ll pay a premium, according to a new study.

The analysis focuses on houses that are likely to be inundated by sea level rise in the year 2100. It suggests that such houses sell for higher prices in neighborhoods inhabited by a relatively large percentage of people who reject the scientific consensus on climate change than they do in neighborhoods inhabited by people who accept climate science.

A growing body of research investigates how climate risk affects asset prices, including real estate. The new study is among the first to explore how beliefs about climate risk affect the real estate market, and the role of future risk in shaping current home prices.

I spent last week at Web Summit in Lisbon, where one of the key takeaways from the event was the extent to which, at many companies, machine learning has moved well beyond proof-of-concepts to being deployed across the business in ways that are having massive, real-world impacts.

Werner Vogels, Amazon’s chief technology officer, gave a keynote detailing how machine learning underpins absolutely everything that the Everything Store does, from recommending products to figuring out which to stock in each fulfillment center to safely operating its new delivery drones.

$106 million initiative will accelerate neuroscience research by embracing artificial intelligence, engineering, data science, other ‘nontraditional’ fields

Gift brings Weill Family Foundation philanthropic giving in neuroscience to over $300 million, enabling bold approaches to curing these diseases

For the first few decades of his career, Peter Turchin had used sophisticated maths to show how the interactions of predators and prey produce oscillations in animal populations in the wild. He had published in the journals Nature and Science and become respected in his field, but by the late 1990s he had answered all the ecological questions that interested him. He found himself drawn to history instead: could the rise and fall of human societies also be captured by a handful of variables and some differential equations?

Turchin set out to determine whether history, like physics, follows certain laws. In 2003, he published a book called Historical Dynamics, in which he discerned secular cycles in France and Russia from their origins to the end of the 18th century. That same year, he founded a new field of academic study, called cliodynamics, which seeks to discover the underlying reasons for these historical patterns, and to model them using mathematics, the way one might model changes to the planet’s climate.

This year, I helped found the Crisis Technologies Innovation Lab at Indiana University, specifically to harness the power of data, technology and artificial intelligence to respond to and prepare for the impacts of climate change.

By analyzing historical disaster information, publicly available census data and predictive models of risk and resilience, our tools will be able to identify and prioritize key decisions, like what kinds of infrastructure investments to make.

The collaboration is significant to the Syracuse community because Microsoft plans to work with the additional partners across the region on devising a broad curriculum of technology and digital literacy programs that can be accessed by local non-profits, community centers, educational institutions, employment and workforce development organizations and businesses. The goal is to more rapidly advance the Syracuse Surge, the community’s strategy of inclusive growth in the new tech economy. The Surge encompasses efforts to transition the region’s workforce, businesses, and innovation centers to address the realities of a tech-based economy. Its goal is to ready workers, businesses, governments, and community organizations for technical change, and to position the region as a leading-edge player in attracting businesses in the new technology space of big data, machine learning, and artificial intelligence.

At the heart of the development is a data set PNNL created in collaboration with Stanford University of over 300,000 medical concepts defined by SNOMED Clinical Terms, a collection of standard medical terms, codes, synonyms and definitions used by medical researchers and practitioners. PNNL developed a graph-based learning method grounded on these terms that outperformed current models. The code is available as an open-source download.

“If you think it’s hard translating doctors’ handwriting, try translating their medical knowledge into computer speak,” observes Robert Rallo, a computer scientist at PNNL who leads the PNNL team applying artificial intelligence to health care. “The tough part is combining multiple types of data. Computer-friendly data like blood work numbers or diagnosis codes are easier than unstructured data like chart notes or images from X-rays or MRIs.”

Google said it has fired an employee for leaking staffer names and personal details to the media and placed two others on leave for allegedly violating company policies, evidence of escalating tension between management and personnel engaged in labor-related activism.

A Google spokeswoman said the company is investigating the employees who were placed on leave. One of them had searched for and shared confidential documents outside the scope of their job, while the other tracked the individual calendars of staff working in the community platforms, human resources, and communications teams, she said. The tracking had made the staff in those departments feel unsafe, the spokeswoman said.

Can scientists fill the void in science journalism? A new study posted on bioRxiv asked this exact question, and found that in terms of article engagement, scientists and journalists engage audiences at roughly equivalent rates. The researchers, led by PhD student Yael Baren-Ben David from the Technion-Israel Institute of Technology, looked at views, clicks, comments and time spent on the page as metrics of engagement, and compared equivalent articles written by scientists and professional journalists. For the two major Israeli online news sites that they studied, the audiences literally and figuratively “liked” articles equivalently no matter who wrote them.

In 2018 $273bn was spent on digital ads globally. We delve into the world of clicks, banners and keywords to find out if any of it is real. What do we really know about the effectiveness of digital advertising?

“In the current era of cybersecurity, your life is part of the attack surface. In this post, we lay out 16 practical steps you can take to secure your data, accounts, and devices. The list is prioritized by risk reduction, so start at #1 and work your way down.”