Cell journal; Jennifer E. Rood, Tim Stuart, Shila Ghazanfar, Tommaso Biancalani, Eyal Fisher, Andrew Butler, Anna Hupalowska, Leslie Gaffney, William Mauck, Gökçen Eraslan, John C. Marioni, Aviv Regev, Rahul Satija
from
Understanding the genetic and molecular drivers of phenotypic heterogeneity across individuals is central to biology. As new technologies enable fine-grained and spatially resolved molecular profiling, we need new computational approaches to integrate data from the same organ across different individuals into a consistent reference and to construct maps of molecular and cellular organization at histological and anatomical scales. Here, we review previous efforts and discuss challenges involved in establishing such a common coordinate framework, the underlying map of tissues and organs. We focus on strategies to handle anatomical variation across individuals and highlight the need for new technologies and analytical methods spanning multiple hierarchical scales of spatial resolution. [full text]
The emergence of artificial intelligence (AI) in health care offers unprecedented opportunities to improve patient and clinical team outcomes, reduce costs, and impact population health. However, it is imperative to balance the need for thoughtful, inclusive health care AI with potential unintended consequences, marketing hype, profit motives, potential user disillusionment, and exacerbation of existing health- and technology-driven disparities.
This new National Academy of Medicine Special Publication synthesizes current knowledge about the use of AI in health care, outlines the current and near-term solutions, highlights challenges and best practices, and identifies limitations of the technology. This Special Publication is viewed as a reference document for all stakeholders involved in AI, health care, or the intersection of the two, as we all move toward a future increasingly dependent on technology together. The Special Publication prioritizes caution in implementation of this technology, prioritization of human connections between clinicians and patients, and an unwavering focus on equity and inclusion.
Many human languages have words for emotions such as “anger” and “fear,” yet it is not clear whether these emotions have similar meanings across languages, or why their meanings might vary. We estimate emotion semantics across a sample of 2474 spoken languages using “colexification”—a phenomenon in which languages name semantically related concepts with the same word. Analyses show significant variation in networks of emotion concept colexification, which is predicted by the geographic proximity of language families. We also find evidence of universal structure in emotion colexification networks, with all families differentiating emotions primarily on the basis of hedonic valence and physiological activation. Our findings contribute to debates about universality and diversity in how humans understand and experience emotion. [full text]
Chinese-language internet search provider Baidu, Inc., and Samsung Electronics Co., Ltd., have announced that Baidu’s first cloud-to-edge AI accelerator, Baidu KUNLUN, has completed its development and will be mass-produced early next year.
Baidu KUNLUN chip is built on the company’s advanced XPU, a home-grown neural processor architecture for cloud, edge, and AI, as well as Samsung’s 14-nanometer (nm) process technology with its I-Cube (Interposer-Cube) package solution. The chip offers 512 gigabytes per second (GBps) memory bandwidth and supplies up to 260 Tera operations per second (TOPS) at 150 watts. In addition, the new chip allows Ernie, a pre-training model for natural language processing, to inference at a three times faster rate than that of conventional GPU/FPGA-accelerating models.
What is sauce for the goose is sauce for the gander when it comes to implementing artificial intelligence in the capital markets industry insiders testified before the House Financial Services Committee’s Taskforce on Artificial Intelligence.
Taskforce chairman and ranking member Rep Bill Foster (IL-D) and Rep. Barry Loudermilk (GA-R) voiced bi-partisan concern that AI, and the big data which fuels it, could upend the capital markets eventually and lead to ever-consolidating markets.
Rep. Foster noted that consolidation and economies-of-scale are the natural byproducts of digital markets, and access to more considerable amounts of data provides fertile soil for market manipulation.
Computers learn to talk (i.e. language models like Bert and specially GPT-2 get scaringly good)
AI becoming good at creating synthetic content has some serious consequences
The biggest theoretical controversy continues to be how to incorporate innate knowledge or structure into machine learned models. There has been little practical progress towards this end, and little progress towards any other theoretical breakthrough.
American Bar Association; Vincent C. Thomas, Justin P. Duda, Travis G. Maurer
from
How does a social media company create and fund a truly independent body to weigh the sometimes competing interests inherent in public discourse?
Facebook’s rather creative answer is to capitalize on Delaware’s policy of freedom of contract and craft a noncharitable purpose trust.
The flexibility of Delaware trust law and the Delaware LLC Act has allowed Facebook to carry out its purpose to provide additional transparency and clarity to its users with respect to content decisions and policies.
University of Virginia (UVA) has received a grant to broaden Virginia universities’ access to protected data for research.
The National Science Foundation gave a $2.5 million grant to the UVA to fund the Virginia Assuring Controls Compliance of Research Data, a collaborative that it runs in coordination with its College at Wise.
The collaborative is run by nearly 11 researchers from Virginia universities to build a high-performance computer system for hosting research using data that by law must be protected. It is the first initiative of its kind in the country to address growing disparities in universities’ access to protected data for research.
The University of Connecticut is aiming to increase commercialization of research, support startups and increase entrepreneurship under a plan that includes a multi-million dollar faculty recruitment effort and new programs at the Stamford campus.
In a draft strategic plan and a proposal for an “Academic Entrepreneurship” initiative, both obtained through a public records request, officials outlined their strategies to turn UConn into a leading institution in “technological innovation and entrepreneurship for economic and social benefit.”
“One of the biggest challenges in slave studies is this idea that people were unknowable, that the slave trade destroyed individuality,” says Daryle Williams, a historian at the University of Maryland. “But the slave trade didn’t erase people. We have all kinds of information that’s knowable—property records, records related to births, deaths and marriages. There are billions of records. It just takes a lot of time to go look at them, and to trace the arc of an individual life.”
In the grand scheme of things, ten years is nothing; an insignificant slice of our planet’s long timeline. But when it comes to technological innovation, a lot can change in a decade, and some of the gadgets, software, and silicon you relied on back in 2010 now seems almost ancient and ready for the antique market.
So let’s reflect on just how far we’ve come, technologically speaking.
West Lafayette, IN May 11-13, 2020. “The symposium will investigate the application of data science in the study of microbiomes and how experimental data and computational data can be leveraged to learn more about microbiome systems.” [save the date]
“The Djokovic Science and Innovation Fellowship supports the research of Harvard University advanced doctoral students whose work is related to early childhood health, learning, and behavior. Selected fellows receive a stipend and join an active, interdisciplinary learning community for one academic year. The goal of the fellowship is to create a new generation of leaders who will drive innovation that impacts the early childhood field and the lives of children facing adversity.” Deadline for applications is January 3, 2020.
Seattle, WA March 23-27, 2020. “Waterhackweek is a 5-day hackweek to be held at the University of Washington in partnership with the University of Washington eScience Institute. Participants will learn about open source technologies used to analyze water-related datasets. Mornings will consist of interactive lectures, and afternoon sessions will involve facilitated exploration of datasets and hands-on software development.” Deadline to apply is January 5, 2020.
“With the objective of improving graduate training, the purpose of this Dear Colleague Letter (DCL) is to draw the attention of the SBE community to the following funding opportunities in the Directorate for Education & Human Resources (EHR) and SBE.”
“The Chan Zuckerberg Initiative invites applications for open source software projects that are essential to biomedical research. Applicants can request funding between $50k and $250k for one year. This RFA is the second of a series.” Deadline for applications is February 4, 2020.
“The workshop will bring together a group of advanced graduate students and a small faculty for an intensive two-week study of computational social science modeling and complexity.” Deadline to apply is February 11, 2020.
“SciPy 2020 will continue the tradition of offering scholarships to attend the conference. These scholarships can provide funding for airfare, lodging, and conference registration.” Deadline to apply is March 19, 2020.
“MIT Hacking Medicine is back with our annual flagship event in Boston, the MIT Grand Hack 2020 on April 17-19. Tackle healthcare’s toughest pain points in healthcare by building innovative solutions with hundreds of like-minded engineers, clinicians, designers, developers and business people. Participants of any background with a passion for healthcare are welcome – with our multi-theme event, there is sure to be a healthcare challenge for everyone!” Deadline to apply is April 8, 2020.
Auto-tinder was created to train an API using Tensorflow and Python3 that learns your interests and automatically plays the tinder swiping-game for you.
University of Maryland, National Socio-Environmental Synthesis Center (SESYNC), Lauren White and Alaina Gallagher
from
“As a scientist, you are a professional writer,” said Dr. Joshua Schimel during a recent science writing workshop held at the National Socio-Environmental Synthesis Center (SESYNC). During this event, SESYNC postdocs had the opportunity to immerse themselves in the storytelling process.
Schimel, a professor at the University of California Santa Barbara, is the author of Writing Science: How to write papers that get cited and proposals that get funded. Using this book as a basis for the workshop, Schimel provided SESYNC postdocs with insights into how to effectively connect with different audiences when writing about their research.