My friend Robert Burton, a neurologist and author, wanted to share a song with me last year, and sent me a link to an NPR Tiny Desk Concert. "It's wonderful to see truly new and inspiring music," he wrote. I clicked open the link to a band who appeared to have journeyed from their mountain village in Russia to busk for tourists in the city square. Three women wore long white wedding dresses, thick strands of bead necklaces, and Cossack hats that towered from their heads like minarets of black wool. They played, respectively, a cello, djembe drum, and floor tom drum. They were joined by an accordion player who could pass for a bearded hipster from Brooklyn. The accordionist was the first to sing. A bray of syllables erupted from him like an exorcism. A steady drumbeat followed and then the women commanded the singing. Their vocals ranged from yodels to yips, whoops to whispers. At first turbulence reigned, as if the women were singing different songs at each other. But soon their voices blended into a melody that curled like a river.

Left ventricular assist devices (LVADs) are mechanical pumps, which can be used to support heart failure (HF) patients as bridge to transplant and destination therapy. To automatically adjust the LVAD speed, a physiological control system needs to be designed to respond to variations of patient hemodynamics across a variety of clinical scenarios. These control systems require pressure feedback signals from the cardiovascular system. However, there are no suitable long-term implantable sensors available. In this study, a novel real-time deep convolutional neural network (CNN) for estimation of preload based on the LVAD flow was proposed. A new sensorless adaptive physiological control system for an LVAD pump was developed using the full dynamic form of model free adaptive control (FFDL-MFAC) and the proposed preload estimator to maintain the patient conditions in safe physiological ranges. The CNN model for preload estimation was trained and evaluated through 10-fold cross validation on 100 different patient conditions and the proposed sensorless control system was assessed on a new testing set of 30 different patient conditions across six different patient scenarios. The proposed preload estimator was extremely accurate with a correlation coefficient of 0.97, root mean squared error of 0.84 mmHg, reproducibility coefficient of 1.56 mmHg, coefficient of variation of 14.44 %, and bias of 0.29 mmHg for the testing dataset. The results also indicate that the proposed sensorless physiological controller works similarly to the preload-based physiological control system for LVAD using measured preload to prevent ventricular suction and pulmonary congestion. This study shows that the LVADs can respond appropriately to changing patient states and physiological demands without the need for additional pressure or flow measurements.

Effective conservation actions require effective population monitoring. However, accurately counting animals in the wild to inform conservation decision-making is difficult. Monitoring populations through image sampling has made data collection cheaper, wide-reaching and less intrusive but created a need to process and analyse this data efficiently. Counting animals from such data is challenging, particularly when densely packed in noisy images. Attempting this manually is slow and expensive, while traditional computer vision methods are limited in their generalisability. Deep learning is the state-of-the-art method for many computer vision tasks, but it has yet to be properly explored to count animals. To this end, we employ deep learning, with a density-based regression approach, to count fish in low-resolution sonar images. We introduce a large dataset of sonar videos, deployed to record wild mullet schools (Mugil liza), with a subset of 500 labelled images. We utilise abundant unlabelled data in a self-supervised task to improve the supervised counting task. For the first time in this context, by introducing uncertainty quantification, we improve model training and provide an accompanying measure of prediction uncertainty for more informed biological decision-making. Finally, we demonstrate the generalisability of our proposed counting framework through testing it on a recent benchmark dataset of high-resolution annotated underwater images from varying habitats (DeepFish). From experiments on both contrasting datasets, we demonstrate our network outperforms the few other deep learning models implemented for solving this task. By providing an open-source framework along with training data, our study puts forth an efficient deep learning template for crowd counting aquatic animals thereby contributing effective methods to assess natural populations from the ever-increasing visual data.

This article presents the data used to evaluate the performance of evolutionary clustering algorithm star (ECA*) compared to five traditional and modern clustering algorithms. Two experimental methods are employed to examine the performance of ECA* against genetic algorithm for clustering++ (GENCLUST++), learning vector quantisation (LVQ) , expectation maximisation (EM) , K-means++ (KM++) and K-means (KM). These algorithms are applied to 32 heterogenous and multi-featured datasets to determine which one performs well on the three tests. For one, ther paper examines the efficiency of ECA* in contradiction of its corresponding algorithms using clustering evaluation measures. These validation criteria are objective function and cluster quality measures. For another, it suggests a performance rating framework to measurethe the performance sensitivity of these algorithms on varos dataset features (cluster dimensionality, number of clusters, cluster overlap, cluster shape and cluster structure). The contributions of these experiments are two-folds: (i) ECA* exceeds its counterpart aloriths in ability to find out the right cluster number; (ii) ECA* is less sensitive towards dataset features compared to its competitive techniques. Nonetheless, the results of the experiments performed demonstrate some limitations in the ECA*: (i) ECA* is not fully applied based on the premise that no prior knowledge exists; (ii) Adapting and utilising ECA* on several real applications has not been achieved yet.

AI has the potential to revolutionize many areas of healthcare. Radiology, dermatology, and ophthalmology are some of the areas most likely to be impacted in the near future, and they have received significant attention from the broader research community. But AI techniques are now also starting to be used in in vitro fertilization (IVF), in particular for selecting which embryos to transfer to the woman. The contribution of AI to IVF is potentially significant, but must be done carefully and transparently, as the ethical issues are significant, in part because this field involves creating new people. We first give a brief introduction to IVF and review the use of AI for embryo selection. We discuss concerns with the interpretation of the reported results from scientific and practical perspectives. We then consider the broader ethical issues involved. We discuss in detail the problems that result from the use of black-box methods in this context and advocate strongly for the use of interpretable models. Importantly, there have been no published trials of clinical effectiveness, a problem in both the AI and IVF communities, and we therefore argue that clinical implementation at this point would be premature. Finally, we discuss ways for the broader AI community to become involved to ensure scientifically sound and ethically responsible development of AI in IVF.

Social norms characterize collective and acceptable group conducts in human society. Furthermore, some social norms emerge from interactions of agents or humans. To achieve agent autonomy and make norm satisfaction explainable, we include emotions into the normative reasoning process, which evaluate whether to comply or violate a norm. Specifically, before selecting an action to execute, an agent observes the environment and infer the state and consequences with its internal states after norm satisfaction or violation of a social norm. Both norm satisfaction and violation provoke further emotions, and the subsequent emotions affect norm enforcement. This paper investigates how modeling emotions affect the emergence and robustness of social norms via social simulation experiments. We find that an ability in agents to consider emotional responses to the outcomes of norm satisfaction and violation (1) promote norm compliance; and (2) improve societal welfare.

Complex systems, such as Artificial Intelligence (AI) systems, are comprised of many interrelated components. In order to represent these systems, demonstrating the relations between components is essential. Perhaps because of this, diagrams, as "icons of relation", are a prevalent medium for signifying complex systems. Diagrams used to communicate AI system architectures are currently extremely varied. The diversity in diagrammatic conceptual modelling choices provides an opportunity to gain insight into the aspects which are being prioritised for communication. In this philosophical exploration of AI systems diagrams, we integrate theories of conceptual models, communication theory, and semiotics. We discuss consequences of standardised diagrammatic languages for AI systems, concluding that while we expect engineers implementing systems to benefit from standards, researchers would have a larger benefit from guidelines.

As Deep Learning continues to drive a variety of applications in datacenters and HPC, there is a growing trend towards building large accelerators with several sub-accelerator cores/chiplets. This work looks at the problem of supporting multi-tenancy on such accelerators. In particular, we focus on the problem of mapping layers from several DNNs simultaneously on an accelerator. Given the extremely large search space, we formulate the search as an optimization problem and develop a specialized genetic algorithm called G# withcustom operators to enable structured sample-efficient exploration. We quantitatively compare G# with several common heuristics, state-of-the-art optimization methods, and reinforcement learning methods across different accelerator set-tings (large/small accelerators) and different sub-accelerator configurations (homogeneous/heterogeneous), and observeG# can consistently find better solutions. Further, to enable real-time scheduling, we also demonstrate a method to generalize the learnt schedules and transfer them to the next batch of jobs, reducing schedule compute time to near zero.

The Covid-19 pandemic has been challenging and, at the same time, a transformative experience for banks and credit unions. While banking providers have invested in digital capabilities, the pandemic pushed them to rethink, redesign and re-engineer their business operations, services and products. To ensure business continuity, financial institutions transitioned to remote working models of sales, digital outreach and customer support. Apart from drastic operations changes, the pandemic also brought along a fundamental change in consumer behavior. Through social distancing and work from home, the average consumer's life was confined to home.

SCI COMMUN### Climate change The United States submitted its new goals to the Paris climate agreement last week, pledging to cut its planet-warming greenhouse gas emissions 50% to 52% below 2005 levels by 2030. The nonbinding pledge, made on Earth Day at an online climate summit convened by President Joe Biden, is one of the most aggressive targets of any wealthy country. The cuts are also higher than the 26% to 28% reduction by 2025 pledged by former President Barack Obama when the United States first joined the agreement in 2015. In the past month, Canada, Japan, and the United Kingdom have also committed to cuts steeper than their earlier pledges; other large polluters, such as China, India, and Russia, have yet to increase their goals ahead of a critical U.N. climate meeting this winter in Glasgow, U.K. The U.S. pledge has a long road to reality. It will require immediate increases in renewable energy, widespread adoption of electric vehicles, and other steps, many of which will require laws that could be difficult to pass, such as Biden's climate-focused infrastructure bill. > “It is our policy … not to employ anyone who has taken the experimental COVID-19 injection.” > > Miami-based private school Centner Academy in a letter to parents this week, citing discredited claims that vaccinated people can transmit harmful substances to others. ### Biomedicine ![Figure][1] CREDITS: (GRAPHIC) J. BRAINARD/ SCIENCE ; (DATA) ASGCT/INFORMA The number of new clinical trials for gene, cell, and RNA therapies has almost tripled in the past 4 years, the American Society of Gene & Cell Therapy said this month in its first ever quarterly trends report. As of 31 March, nearly 3500 of these experimental treatments were in development, most of which (53%) consist of altered genes or genetically modified (GM) cells, such as cancer-fighting T cells with modified receptors. Cancer is the target for the largest number (1200) of therapies under development. Among RNA therapies, many were vaccines (35) or COVID-19 treatments (30). The United States has more clinical trials underway in each of the three types of therapies than any other country: 1400 overall. Globally, 16 gene therapies (including GM cells), 53 non-GM cell therapies, and 15 RNA therapies have been approved for use so far. ### Vaccination The Department of Health and Human Services (HHS) last week killed a rule that would have made it much more difficult for people who sustain shoulder injuries during vaccination to win compensation from a $4.1 billion government fund. The rule had been finalized on 19 January, the last day of former President Donald Trump's administration. But President Joe Biden's administration froze its implementation on 20 January. The new administration said the previous one had been “irregular in its haste” when it moved to remove shoulder injuries from a list of injuries in which the petitioner does not have to prove a vaccine caused the injury, making it easier to win a government payout ( Science , 10 April 2020, p. [121][2]). Shoulder injuries accounted for nearly 55% of more than 2400 claims filed with the National Vaccine Injury Compensation Program in the past 2 years, most of them after flu shots. COVID-19 vaccines fall under a different HHS program, but if they win formal approval by the U.S. Food and Drug Administration, they could in principle be added to the national compensation program. ### Publishing Sixteen journals, including BMJ Open Science and Royal Society Open Science , say they will accept articles reviewed by the nonprofit Peer Community In Registered Reports (PCI RR). The organization, launched last week, will review one type of article: “registered reports,” describing studies for which detailed experimental plans are peer reviewed before research begins. Once the research is complete, PCI RR will do a second round of peer review including results and analysis. Papers it recommends can then be published in any of the 16 journals without further review as long as they meet a journal's normal criteria. The organization, funded by donations, will provide peer review free to authors and journals in any discipline. ### COVID-19 The National Institutes of Health (NIH) last week announced it would launch a large study repurposing existing drugs for patients with mild COVID-19 symptoms who don't need hospitalization. The $155 million trial, which aims to recruit 13,500 participants, will open in a few weeks at several research centers. It will test up to seven medications already approved for other conditions, but NIH hasn't yet named them. A recent, similar trial at the University of Oxford that has so far enrolled about 4800 patients has found that the asthma drug budesonide lessens symptoms and speeds recovery in certain COVID-19 patients. ### Medicine A large study has firmed up earlier evidence that SARS-CoV-2 increases the rate of complications for pregnant women and their babies. The study followed 706 pregnant women with COVID-19 and 1424 uninfected pregnant women at hospitals in 18 countries. Infected women had a 76% higher risk of developing problems caused by pregnancy-associated high blood pressure and a 59% higher risk of preterm birth. They were also five times more likely to be admitted to intensive care than uninfected women. Eleven women with COVID-19 died, compared with one uninfected woman, researchers report in JAMA Pediatrics . Infected women with fever and shortness of breath had babies with a fivefold increased risk of complications such as immature lungs, eye disorders, and brain damage. The coronavirus may affect pregnancy via changes in a woman's heart, lungs, and immune system. The results show pregnant women should be among priority groups for COVID-19 vaccines, the authors say. A separate study last week found no obvious safety problems in more than 800 U.S. women who gave birth after receiving messenger RNA vaccines. ### Community The American Humanist Association has decided to withdraw Richard Dawkins's 1996 Humanist of the Year award as a result of his “history of making statements that use the guise of scientific discourse to demean marginalized groups.” The decision came soon after the evolutionary biologist and former University of Oxford professor wrote a tweet comparing transgender people to Rachel Dolezal, a civil rights activist who for years posed as Black. The association said the tweet “implies that the identities of transgender individuals are fraudulent, while also simultaneously attacking Black identity as one that can be assumed when convenient.” In 2015, Dawkins argued that trans women are not women based on their chromosomes, but said he would use the pronoun “she” out of courtesy. ### Seismology The first large-scale, phone-based earthquake early warning system will be deployed in Greece and New Zealand, Google announced this week. Since last year, the company has been testing the use of data compiled from its more than 2 billion active Android phones to pinpoint the location and strength of earthquakes. The measurement comes from the built-in Android phones' accelerometers, which sense movement just like seismometers. When the phones detect earthquakelike signals, they alert a server that combines information from many phones. If enough phones corroborate the result, an alert goes out. The cellphone results compared well to those from seismometer-based warning systems in Japan and the United States. Google chose New Zealand and Greece, both countries with high earthquake hazards and many Android phones, to premier the system because they lack operational warning systems of their own. Eventually, phone-based alerts could be available worldwide. ### Biotechnology Genetically modified mosquitoes designed to prevent the spread of viruses such as Zika and dengue are set to be released in the United States for the first time. Starting this week, the company Oxitec will free fewer than 12,000 transgenic Aedes aegypti mosquitoes, all of them nonbiting males, in the Florida Keys as part of a pilot study, the company announced on 23 April. They are engineered to carry a gene that kills their female offspring, reducing the population of mosquitoes capable of transmitting diseases. Field tests outside the United States have shown dramatic population drops, though the company has not published definitive evidence that the strategy reduces disease in humans. The project, which has long faced public opposition in Florida, won approval in May 2020 from the Environmental Protection Agency, which predicted no adverse effects on people or local wildlife. ### Science policy President Joe Biden last week picked two veterans of government service and a newcomer to fill top science positions. He named soil scientist Asmeret Asefaw Berhe of the University of California, Merced, to lead the Department of Energy's Office of Science. Berhe, born in Eritrea, has little government experience but has won accolades for her research and efforts to promote diversity in science. She would become the first Black woman to lead the science office if confirmed by the Senate. Oceanographer Rick Spinrad of Oregon State University, Corvallis, who has held numerous posts at the National Oceanic and Atmospheric Administration, is Biden's choice to lead that agency. To run the State Department's science bureau, he chose Monica Medina, an ocean policy expert and attorney at Georgetown University. Confirmation hearings for the three nominees could come as early as next month. ### Regulation In one of the broadest attempts to regulate artificial intelligence (AI) to date, the European Commission on 21 April proposed new rules for algorithms that power everything from medical device and credit scoring software to chatbots and facial recognition systems. The rules divide AI technologies into risk categories, and put outright bans on some, such as systems that would score individuals' “social credit.” Other “high-risk” systems, including those that collect biometric data, would require a strict vetting process. However, there are exceptions for national security, and it could take years before the rules become law: They must first pass the European Council and the European Parliament and be adopted by member countries. ### Community An effort to increase gender, racial, and geographic diversity in the U.S. National Academy of Sciences (NAS) has begun to bear fruit. Nearly half of the 120-member 2021 class announced last week are women, compared with one-quarter in 2011. The new cohort includes nine Black scientists; NAS officials say previous classes never had more than three and often had none. “We need to do better, but I'm amazed at how far we've come,” says plant geneticist Susan Wessler, NAS home secretary. The academy's governing council has begun to give more slots to disciplinary units that present candidate slates less skewed toward older white men. To reduce its geographic imbalance, NAS is also prohibiting members from nominating someone from their own institution. Today, 18 U.S. states have two or fewer members, whereas a handful of elite academic institutions each have more than 100. [1]: pending:yes [2]: http://www.sciencemag.org/content/368/6487/121