SYDNEY – BGI Group, the world's largest genomics company, has worked with China's military on research that ranges from mass testing for respiratory pathogens to brain science, a review of research, patent filings and other documents has found. The review, of more than 40 publicly available documents and research papers in Chinese and English, shows BGI's links to the People's Liberation Army (PLA) include research with China's top military supercomputing experts. The extent of those links has not previously been reported. BGI has sold millions of COVID-19 test kits outside China since the outbreak of the new coronavirus pandemic, including to Europe, Australia and the United States. Shares of BGI Genomics Co., the company's subsidiary listed on the Shenzhen stock exchange, have doubled in price over the past 12 months, giving it a market value of about $9 billion.

Autonomous Vehicles (AVs) raise important social and ethical concerns, especially about accountability, dignity, and justice. We focus on the specific concerns arising from how AV technology will affect the lives and livelihoods of professional and semi-professional drivers. Whereas previous studies of such concerns have focused on the opinions of experts, we seek to understand these ethical and societal challenges from the perspectives of the drivers themselves. To this end, we adopted a qualitative research methodology based on semi-structured interviews. This is an established social science methodology that helps understand the core concerns of stakeholders in depth by avoiding the biases of superficial methods such as surveys. We find that whereas drivers agree with the experts that AVs will significantly impact transportation systems, they are apprehensive about the prospects for their livelihoods and dismiss the suggestions that driving jobs are unsatisfying and their profession does not merit protection. By showing how drivers differ from the experts, our study has ramifications beyond AVs to AI and other advanced technologies. Our findings suggest that qualitative research applied to the relevant, especially disempowered, stakeholders is essential to ensuring that new technologies are introduced ethically.

Fox News Flash top entertainment and celebrity headlines are here. Check out what's clicking today in entertainment. Cloris Leachman, known for her decades-long career in film, television and beyond, has died at the age of 94. She died on Wednesday of natural causes, her rep told Fox News. Leachman was a history-making actress, having racked up more Emmy award wins than any other performer in the business with eight awards for primetime programming and an additional Daytime Emmy for appearing in "ABC Afterschool Specials."

Care and Feeding is Slate's parenting advice column. Have a question for Care and Feeding? Submit it here or post it in the Slate Parenting Facebook group. My husband and I both have full-time jobs and an 18-month-old son. I am pregnant with our second child, due in February. Since our son was born, my husband seems to have regressed.

My father, Bill Broderick, who has died aged 80 of Covid-19, was an educationist ahead of his time in the field of computing. His vision and enthusiasm led to the first computer being installed in a British secondary school, the Royal Liberty school in Romford, Essex, where he was a maths teacher, in 1965. In a broadcast by the BBC programme Tomorrow's World from the school, Bill said: "Computers are as radical and important a keystone to our standard of living and industrial wellbeing as was the steam engine." Born in Farnborough, Kent, Bill was the only son of Ralph Broderick, an engineer, and Ida (nee Massey). He was educated at Lord Wandsworth college in Long Sutton, Hampshire, then went to Hull University to study mathematics.

Although deep reinforcement learning has led to breakthroughs in many difficult domains, these successes have required an ever-increasing number of samples, affording only a shrinking segment of the AI community access to their development. Resolution of these limitations requires new, sample-efficient methods. To facilitate research in this direction, we propose this second iteration of the MineRL Competition. The primary goal of the competition is to foster the development of algorithms which can efficiently leverage human demonstrations to drastically reduce the number of samples needed to solve complex, hierarchical, and sparse environments. To that end, participants compete under a limited environment sample-complexity budget to develop systems which solve the MineRL ObtainDiamond task in Minecraft, a sequential decision making environment requiring long-term planning, hierarchical control, and efficient exploration methods. The competition is structured into two rounds in which competitors are provided several paired versions of the dataset and environment with different game textures and shaders. At the end of each round, competitors submit containerized versions of their learning algorithms to the AIcrowd platform where they are trained from scratch on a hold-out dataset-environment pair for a total of 4-days on a pre-specified hardware platform. In this follow-up iteration to the NeurIPS 2019 MineRL Competition, we implement new features to expand the scale and reach of the competition. In response to the feedback of the previous participants, we introduce a second minor track focusing on solutions without access to environment interactions of any kind except during test-time. Further we aim to prompt domain agnostic submissions by implementing several novel competition mechanics including action-space randomization and desemantization of observations and actions.

Conventional algorithmic fairness is West-centric, as seen in its sub-groups, values, and methods. In this paper, we de-center algorithmic fairness and analyse AI power in India. Based on 36 qualitative interviews and a discourse analysis of algorithmic deployments in India, we find that several assumptions of algorithmic fairness are challenged. We find that in India, data is not always reliable due to socio-economic factors, ML makers appear to follow double standards, and AI evokes unquestioning aspiration. We contend that localising model fairness alone can be window dressing in India, where the distance between models and oppressed communities is large. Instead, we re-imagine algorithmic fairness in India and provide a roadmap to re-contextualise data and models, empower oppressed communities, and enable Fair-ML ecosystems.

Food production is a critical activity in which every nation would like to be self-sufficient. India is one of the largest producers of food grains in the world. In India, nearly 70 percent of rural households still depend on agriculture for their livelihood. Keeping farmers happy is particularly important in India as farmers form a large vote bank which politicians dare not disappoint. At the same time, Governments need to balance the interest of farmers with consumers, intermediaries and society at large. The whole agriculture sector is highly information-intensive. Even with enormous collection of data and statistics from different arms of Government, there continue to be information gaps. In this paper we look at how Tantra Social Information Management Framework can help analyze the agricultural sector and transform the same using a holistic approach. Advantage of Tantra Framework approach is that it looks at societal information as a whole without limiting it to only the sector at hand. Tantra Framework makes use of concepts from Zachman Framework to manage aspects of social information through different perspectives and concepts from Unified Foundational Ontology (UFO) to represent interrelationships between aspects. Further, Tantra Framework interoperates with models such as Balanced Scorecard, Theory of Change and Theory of Separations. Finally, we model Indian Agricultural Sector as a business ecosystem and look at approaches to steer transformation from within.

SCI COMMUN SCIENTISTS' TERM LIMITS The Trump administration moved to impose 5-year term limits on top scientists at the U.S. Department of Health and Human Services (HHS). The rule, released on 15 January, requires that directors of seven centers at the Food and Drug Administration, as well as 17 positions at the Centers for Disease Control and Prevention, undergo a performance review that could lead to a new 5-year appointment, or to the staffer's transfer. A 2016 law mandates such 5-year reviews for institute and center directors at the National Institutes of Health. But some current and former officials worry the term limits will subject such positions to political interference from the White House, and they could face legal challenges, Politico reported. GLOBAL WARMING A surprise Environmental Protection Agency (EPA) rule finalized last week would effectively ban the government from regulating greenhouse gas emissions from heavy industries other than power plants. The agency substantially rewrote a draft rule originally focused on regulating carbon emissions from new power plants, expanding it to exempt other “stationary” sources, such as refineries and oil and gas wells. The exemption covers an entire class of sources if its collective emissions are less than 3% of the U.S. total. Only power plants, which produce 27% of U.S. carbon emissions, exceed that bar. Analysts say the rule is vulnerable to a court challenge, and the Biden administration can likely suspend its implementation. SPOTTED OWLS The U.S. Fish and Wildlife Service slashed protections for the endangered northern spotted owl on 13 January, declaring more than 1.4 million hectares of Pacific Northwest forests would no longer be considered critical habitat for the bird. The decision comes despite findings by agency scientists that the owl's population is declining and that it warrants stricter protection. The habitat reduction is part of a move by the outgoing Trump administration to settle a lawsuit by the timber industry and counties that earn revenue from logging. The land area is 17 times the amount that the agency initially proposed to remove from protections in August 2020. TRANSGENIC ANIMALS A push by the White House would essentially eliminate the Food and Drug Administration's (FDA's) authority to regulate genetically modified animals and put the U.S. Department of Agriculture (USDA) in charge. The two agencies had been negotiating on dividing the task, and critics of the White House move say it would put USDA in the problematic position of both promoting and regulating genetically modified animals. FDA opposes the shift, Politico reported. ARCTIC OIL DRILLING The first-ever auction of oil drilling rights inside the Arctic National Wildlife Refuge in Alaska, a policy priority for the Trump administration, met with a tepid response this month. Just three bidders paid $14.4 million to claim 11 parcels covering 220,000 hectares—about half of the land up for auction. In an unusual move, a state agency, the Alaska Industrial Development and Export Authority, won bids for nine of the parcels. It joined the auction in part because the agency fears the Biden administration will slow or block further leasing, reducing the state's potential economic gains. The leases must still be finalized by the Bureau of Land Management. Wildlife scientists have warned that the drilling could harm caribou herds and other parts of the ecosystem. FETAL TISSUE RESEARCH Scientists who use fetal tissue obtained from elective abortions would need to comply with new rules under a proposal released by the Trump administration on 13 January. Among other changes, the policy would add new requirements to forms used to obtain informed consent from women who donate tissue for research. It would also limit the source of fetal tissue, which often comes from nonprofit clinics, to federally or state funded hospitals or academic medical centers. In 2019, Trump's administration banned fetal tissue research by federal researchers and required a new ethics review for studies by scientists receiving federal grants; research groups have urged the Biden administration to reverse that policy. The new proposal, which is open for comment for 30 days, is not expected to move forward. CENSUS FIGHT The Trump administration last week abandoned a 2-year effort to prod the Census Bureau to provide a separate tally of undocumented U.S. residents as part of the 2020 census. In 2019, the president had ordered that the separate tally, and a rushed compilation of the decennial head count used for apportioning the 435 seats of the U.S. House of Representatives, be delivered before he left office. Most demographers said it could not be done and called the directive political interference. On 11 January, U.S. government lawyers told a federal judge that the apportionment numbers would not be ready until 6 March. Civil rights organizations want Census Director Steven Dillingham to resign before his term ends in December, saying he has failed to uphold the agency's high standards for data quality. ### Leadership President Joe Biden on 15 January named Eric Lander to be his science adviser and director of the White House Office of Science and Technology Policy. A mathematician turned molecular biologist, the 63-year-old Lander will take leave from his post as president and founding director of the Broad Institute, jointly run by Harvard University and the Massachusetts Institute of Technology (MIT). The first biologist to hold the job, Lander spent 8 years as co-chair of the nation's top science advisory panel under former President Barack Obama. He also co-led the public Human Genome Project, which completed its first draft in 2001. Biden has picked chemistry Nobel laureate Frances Arnold and MIT's Maria Zuber to lead his science advisory panel. He named David Kessler, a former commissioner of the U.S. Food and Drug Administration, to direct Operation Warp Speed, the federal effort to speed development of COVID-19 vaccines. And Biden said Francis Collins has agreed to remain as director of the National Institutes of Health. ### Conservation To study one of Europe's rarest butterflies, researchers pioneered a new method of observation: rappelling down vertiginous mountainsides along the border of Italy and Switzerland. Scientists first described the orange-and-brown Raetzer's ringlet ( Erebia christi ) more than 100 years ago, but its dangerous, inaccessible habitat complicated population surveys. Drawing on decades of climbing experience, independent biologists Andrea Battisti and Matteo Gabaglio slid down ropes to count butterflies in several areas during the past 6 years. It was “like being an explorer … going where nobody has ever [gone],” Gabaglio says. Researchers sighted the ringlet 177 times at two key sites in Italy, they reported this month in the Journal of Insect Conservation . That's good news, they add: The ringlet appears to be more abundant that previous studies suggested. But because of climate change and other threats, they recommend reclassifying the species as endangered rather than vulnerable. ### Policy President Joe Biden announced a sweeping, $400 billion plan last week to tackle the “dismal failure” of the COVID-19 vaccine rollout, safely reopen schools by March, and ramp up testing people for the pandemic coronavirus. The measures are part of an ambitious, $1.9 billion “American Rescue Plan” unveiled by Biden ahead of his inauguration to help people who are struggling financially because of the pandemic—a proposal that depends on Congress providing the money. The federal government would pay for 100,000 new public health workers to assist states in vaccination and other pandemic response efforts. Biden promised to invoke the Defense Production Act to provide vaccinemakers with whatever they need to increase production. Biden's administration would also work more closely with pharmacies to move vaccines from freezers into arms. “The more people we vaccinate and the faster we do it, the sooner we can put this pandemic behind us,” Biden said. ### Infectious diseases All eight gorillas at the San Diego Zoo Safari Park were exposed to the coronavirus that causes COVID-19, and at least two have begun to cough, zoo officials said last week. Tests of fecal samples showed that two were infected, marking the first known cases in nonhuman apes. The officials suspect the western lowland gorillas caught the virus from an asymptomatic staff member who tested positive for SARS-CoV-2; the zoo has been closed to the public for weeks because of the pandemic. The news confirmed fears that the virus can infect endangered great apes. Human respiratory viruses are already a leading cause of death for chimpanzees in the wild. ### COVID-19 High virus levels in saliva are correlated with later hospitalization, serious illness, or death from COVID-19, raising the prospect that testing saliva for the coronavirus that causes the disease will help identify patients most at risk, a study has found. The standard test to detect the SARS-CoV-2 virus analyzes samples of nasal mucus taken with nasopharyngeal (NP) swabs. But patients with the worst outcomes were more likely to have high virus loads in their saliva, but not in their NP swabs, report Akiko Iwasaki of Yale University and colleagues in a 10 January preprint. That may reflect that nasal mucus comes from the upper respiratory tract, whereas severe disease is associated with damage deep in the lungs; coughing regularly brings up viral particles to the throat, where they can pervade saliva. If the results are confirmed, saliva tests could help doctors prioritize which patients in the early stages of the disease should receive medicines that drive down levels of the virus. ### Planetary science The Red Planet has claimed another robot. Scientists at NASA and the German Aerospace Center last week called off a 2-year effort to rescue the failed rod-shaped heat probe, or “mole,” of the InSight lander. The mole was designed to burrow 5 meters into the martian soil and tease out how quickly heat escapes from Mars—a clue to how the planet formed. But soil compacted instead of crumbling as the rod tried to dig in, leaving it stuck at the surface. Even after engineers used InSight's robotic arm to push the probe down and scraped dirt on top, the probe failed a final attempt this month to dig on its own, leaving the mole buried in a shallow grave. InSight's other primary instrument, a seismometer, continues to function normally. ### Foreign influences The U.S. Department of Justice (DOJ) has added another prominent scientist to its crackdown on U.S.-based academics with allegedly undisclosed ties to China. On 14 January, police arrested nanotechnologist Gang Chen of the Massachusetts Institute of Technology (MIT) at his Cambridge home, charging him with violating federal wire fraud, banking, and tax laws. DOJ alleges Chen held various appointments with Chinese institutions and provided technical advice, “often in exchange for financial compensation and awards.” He allegedly failed to disclose these affiliations as required when applying for U.S. Department of Energy grants, and did not tell tax authorities about a bank account in China. MIT said, “We take seriously concerns about improper influence in U.S. research.” Chen was born in China and is a naturalized U.S. citizen. He is a member of the National Academy of Engineering and fellow of AAAS (which publishes Science ). ### History of science Fans of Mary Anning are hoping to raise £33,000 by next month to fund a statue honoring the paleontological pioneer, who discovered and interpreted key fossils along England's Jurassic Coast. Anning, who lived in Lyme Regis in the early 1800s, was the first to correctly identify an ichthyosaur, and discovered England's first pterosaur. Her discoveries were profoundly influential, but as a self-taught, working-class woman she was excluded from meetings of the Geological Society of London. Its members discussed and built on her discoveries, but often failed to acknowledge her. The £100,000 statue project was inspired by 13-year-old local Evie Swire. Organizers hope the cause will be helped by the film Ammonite , starring Kate Winslet as Anning, which opened in U.S. theaters in November 2020. ### Publishing AAAS, which publishes the Science family of journals, said last week it will offer its authors a free way to comply with funder requirements that their papers be open access on publication. Under a new policy, authors may deposit near-final, peer-reviewed versions of papers accepted by paywalled Science titles in public repositories where they are free to read. This “green open-access” route will apply for now only to authors of papers funded by Coalition S, a group of mostly European funders and foundations behind a mandate for immediate open access that takes effect this month. AAAS said it will pilot the new policy for 1 year.

This thesis is a proof of concept for the potential of Variational Auto-Encoder (VAE) on representation learning of real-world Knowledge Graphs (KG). Inspired by successful approaches to the generation of molecular graphs, we evaluate the capabilities of our model, the Relational Graph Variational Auto-Encoder (RGVAE). The impact of the modular hyperparameter choices, encoding through graph convolutions, graph matching and latent space prior, is compared. The RGVAE is first evaluated on link prediction. The mean reciprocal rank (MRR) scores on the two datasets FB15K-237 and WN18RR are compared to the embedding-based model DistMult. A variational DistMult and a RGVAE without latent space prior constraint are implemented as control models. The results show that between different settings, the RGVAE with relaxed latent space, scores highest on both datasets, yet does not outperform the DistMult. Further, we investigate the latent space in a twofold experiment: first, linear interpolation between the latent representation of two triples, then the exploration of each latent dimension in a $95\%$ confidence interval. Both interpolations show that the RGVAE learns to reconstruct the adjacency matrix but fails to disentangle. For the last experiment we introduce a new validation method for the FB15K-237 data set. The relation type-constrains of generated triples are filtered and matched with entity types. The observed rate of valid generated triples is insignificantly higher than the random threshold. All generated and valid triples are unseen. A comparison between different latent space priors, using the $\delta$-VAE method, reveals a decoder collapse. Finally we analyze the limiting factors of our approach compared to molecule generation and propose solutions for the decoder collapse and successful representation learning of multi-relational KGs.