Paper by Takahiro Yabe et al: “Disruptions, such as closures of businesses during pandemics, not only affect businesses and amenities directly but also influence how people move, spreading the impact to other businesses and increasing the overall economic shock. However, it is unclear how much businesses depend on each other during disruptions. Leveraging human mobility data and same-day visits in five US cities, we quantify dependencies between points of interest encompassing businesses, stores and amenities. We find that dependency networks computed from human mobility exhibit significantly higher rates of long-distance connections and biases towards specific pairs of point-of-interest categories. We show that using behaviour-based dependency relationships improves the predictability of business resilience during shocks by around 40% compared with distance-based models, and that neglecting behaviour-based dependencies can lead to underestimation of the spatial cascades of disruptions. Our findings underscore the importance of measuring complex relationships in patterns of human mobility to foster urban economic resilience to shocks…(More)”.
Big brother: the effects of surveillance on fundamental aspects of social vision
Paper by Kiley Seymour et al: “Despite the dramatic rise of surveillance in our societies, only limited research has examined its effects on humans. While most research has focused on voluntary behaviour, no study has examined the effects of surveillance on more fundamental and automatic aspects of human perceptual awareness and cognition. Here, we show that being watched on CCTV markedly impacts a hardwired and involuntary function of human sensory perception—the ability to consciously detect faces. Using the method of continuous flash suppression (CFS), we show that when people are surveilled (N = 24), they are quicker than controls (N = 30) to detect faces. An independent control experiment (N = 42) ruled out an explanation based on demand characteristics and social desirability biases. These findings show that being watched impacts not only consciously controlled behaviours but also unconscious, involuntary visual processing. Our results have implications concerning the impacts of surveillance on basic human cognition as well as public mental health…(More)”.
Data solidarity: Operationalising public value through a digital tool
Paper by Seliem El-Sayed, Ilona Kickbusch & Barbara Prainsack: “Most data governance frameworks are designed to protect the individuals from whom data originates. However, the impacts of digital practices extend to a broader population and are embedded in significant power asymmetries within and across nations. Further, inequities in digital societies impact everyone, not just those directly involved. Addressing these challenges requires an approach which moves beyond individual data control and is grounded in the values of equity and a just contribution of benefits and risks from data use. Solidarity-based data governance (in short: data solidarity), suggests prioritising data uses over data type and proposes that data uses that generate public value should be actively facilitated, those that generate significant risks and harms should be prohibited or strictly regulated, and those that generate private benefits with little or no public value should be ‘taxed’ so that profits generated by corporate data users are reinvested in the public domain. In the context of global health data governance, the public value generated by data use is crucial. This contribution clarifies the meaning, importance, and potential of public value within data solidarity and outlines methods for its operationalisation through the PLUTO tool, specifically designed to assess the public value of data uses…(More)”.
The AI tool that can interpret any spreadsheet instantly
Article by Duncan C. McElfresh: “Say you run a hospital and you want to estimate which patients have the highest risk of deterioration so that your staff can prioritize their care1. You create a spreadsheet in which there is a row for each patient, and columns for relevant attributes, such as age or blood-oxygen level. The final column records whether the person deteriorated during their stay. You can then fit a mathematical model to these data to estimate an incoming patient’s deterioration risk. This is a classic example of tabular machine learning, a technique that uses tables of data to make inferences. This usually involves developing — and training — a bespoke model for each task. Writing in Nature, Hollmann et al.report a model that can perform tabular machine learning on any data set without being trained specifically to do so.
Tabular machine learning shares a rich history with statistics and data science. Its methods are foundational to modern artificial intelligence (AI) systems, including large language models (LLMs), and its influence cannot be overstated. Indeed, many online experiences are shaped by tabular machine-learning models, which recommend products, generate advertisements and moderate social-media content3. Essential industries such as healthcare and finance are also steadily, if cautiously, moving towards increasing their use of AI.
Despite the field’s maturity, Hollmann and colleagues’ advance could be revolutionary. The authors’ contribution is known as a foundation model, which is a general-purpose model that can be used in a range of settings. You might already have encountered foundation models, perhaps unknowingly, through AI tools, such as ChatGPT and Stable Diffusion. These models enable a single tool to offer varied capabilities, including text translation and image generation. So what does a foundation model for tabular machine learning look like?
Let’s return to the hospital example. With spreadsheet in hand, you choose a machine-learning model (such as a neural network) and train the model with your data, using an algorithm that adjusts the model’s parameters to optimize its predictive performance (Fig. 1a). Typically, you would train several such models before selecting one to use — a labour-intensive process that requires considerable time and expertise. And of course, this process must be repeated for each unique task.
In the hands of a few: Disaster recovery committee networks
Paper by Timothy Fraser, Daniel P. Aldrich, Andrew Small and Andrew Littlejohn: “When disaster strikes, urban planners often rely on feedback and guidance from committees of officials, residents, and interest groups when crafting reconstruction policy. Focusing on recovery planning committees after Japan’s 2011 earthquake, tsunami, and nuclear disasters, we compile and analyze a dataset on committee membership patterns across 39 committees with 657 members. Using descriptive statistics and social network analysis, we examine 1) how community representation through membership varied among committees, and 2) in what ways did committees share members, interlinking members from certain interests groups. This study finds that community representation varies considerably among committees, negatively related to the prevalence of experts, bureaucrats, and business interests. Committee membership overlap occurred heavily along geographic boundaries, bridged by engineers and government officials. Engineers and government bureaucrats also tend to be connected to more members of the committee network than community representatives, giving them prized positions to disseminate ideas about best practices in recovery. This study underscores the importance of diversity and community representation in disaster recovery planning to facilitate equal participation, information access, and policy implementation across communities…(More)”.
Survey of attitudes in a Danish public towards reuse of health data
Paper by Lea Skovgaard et al: “Everyday clinical care generates vast amounts of digital data. A broad range of actors are interested in reusing these data for various purposes. Such reuse of health data could support medical research, healthcare planning, technological innovation, and lead to increased financial revenue. Yet, reuse also raises questions about what data subjects think about the use of health data for various different purposes. Based on a survey with 1071 respondents conducted in 2021 in Denmark, this article explores attitudes to health data reuse. Denmark is renowned for its advanced integration of data infrastructures, facilitating data reuse. This is therefore a relevant setting from which to explore public attitudes to reuse, both as authorities around the globe are currently working to facilitate data reuse opportunities, and in the light of the recent agreement on the establishment in 2024 of the European Health Data Space (EHDS) within the European Union (EU). Our study suggests that there are certain forms of health data reuse—namely transnational data sharing, commercial involvement, and use of data as national economic assets—which risk undermining public support for health data reuse. However, some of the purposes that the EHDS is supposed to facilitate are these three controversial purposes. Failure to address these public concerns could well challenge the long-term legitimacy and sustainability of the data infrastructures currently under construction…(More)”
Participatory seascape mapping: A community-based approach to ocean governance and marine conservation
Paper by Isabel James: “Despite the global proliferation of ocean governance frameworks that feature socioeconomic variables, the inclusion of community needs and local ecological knowledge remains underrepresented. Participatory mapping or Participatory GIS (PGIS) has emerged as a vital method to address this gap by engaging communities that are conventionally excluded from ocean planning and marine conservation. Originally developed for forest management and Indigenous land reclamation, the scholarship on PGIS remains predominantly focused on terrestrial landscapes. This review explores recent research that employs the method in the marine realm, detailing common methodologies, data types and applications in governance and conservation. A typology of ocean-centered PGIS studies was identified, comprising three main categories: fisheries, habitat classification and blue economy activities. Marine Protected Area (MPA) design and conflict management are the most prevalent conservation applications of PGIS. Case studies also demonstrate the method’s effectiveness in identifying critical marine habitats such as fish spawning grounds and monitoring endangered megafauna. Participatory mapping shows particular promise in resource and data limited contexts due to its ability to generate large quantities of relatively reliable, quick and low-cost data. Validation steps, including satellite imagery and ground-truthing, suggest encouraging accuracy of PGIS data, despite potential limitations related to human error and spatial resolution. This review concludes that participatory mapping not only enriches scientific research but also fosters trust and cooperation among stakeholders, ultimately contributing to more resilient and equitable ocean governance…(More)”.
Distorted insights from human mobility data
Paper by Riccardo Gallotti, Davide Maniscalco, Marc Barthelemy & Manlio De Domenico: “The description of human mobility is at the core of many fundamental applications ranging from urbanism and transportation to epidemics containment. Data about human movements, once scarce, is now widely available thanks to new sources such as phone call detail records, GPS devices, or Smartphone apps. Nevertheless, it is still common to rely on a single dataset by implicitly assuming that the statistical properties observed are robust regardless of data gathering and processing techniques. Here, we test this assumption on a broad scale by comparing human mobility datasets obtained from 7 different data-sources, tracing 500+ millions individuals in 145 countries. We report wide quantifiable differences in the resulting mobility networks and in the displacement distribution. These variations impact processes taking place on these networks like epidemic spreading. Our results point to the need for disclosing the data processing and, overall, to follow good practices to ensure robust and reproducible results…(More)”
SciAgents: Automating Scientific Discovery Through Bioinspired Multi-Agent Intelligent Graph Reasoning
Paper by Alireza Ghafarollahi, and Markus J. Buehler: “A key challenge in artificial intelligence (AI) is the creation of systems capable of autonomously advancing scientific understanding by exploring novel domains, identifying complex patterns, and uncovering previously unseen connections in vast scientific data. In this work, SciAgents, an approach that leverages three core concepts is presented: (1) large-scale ontological knowledge graphs to organize and interconnect diverse scientific concepts, (2) a suite of large language models (LLMs) and data retrieval tools, and (3) multi-agent systems with in-situ learning capabilities. Applied to biologically inspired materials, SciAgents reveals hidden interdisciplinary relationships that were previously considered unrelated, achieving a scale, precision, and exploratory power that surpasses human research methods. The framework autonomously generates and refines research hypotheses, elucidating underlying mechanisms, design principles, and unexpected material properties. By integrating these capabilities in a modular fashion, the system yields material discoveries, critiques and improves existing hypotheses, retrieves up-to-date data about existing research, and highlights strengths and limitations. This is achieved by harnessing a “swarm of intelligence” similar to biological systems, providing new avenues for discovery. How this model accelerates the development of advanced materials by unlocking Nature’s design principles, resulting in a new biocomposite with enhanced mechanical properties and improved sustainability through energy-efficient production is shown…(More)”.
Protecting civilians in a data-driven and digitalized battlespace: Towards a minimum basic technology infrastructure
Paper by Ann Fitz-Gerald and Jenn Hennebry: “This article examines the realities of modern day warfare, including a rising trend in hybrid threats and irregular warfare which employ emerging technologies supported by digital and data-driven processes. The way in which these technologies become applied generates a widened battlefield and leads to a greater number of civilians being caught up in conflict. Humanitarian groups mandated to protect civilians have adapted their approaches to the use of new emerging technologies. However, the lack of international consensus on the use of data, the public and private nature of the actors involved in conflict, the transnational aspects of the widened battlefield, and the heightened security risks in the conflict space pose enormous challenges for the protection of civilians agenda. Based on the dual-usage aspect of emerging technologies, the challenges associated with regulation and the need for those affected by conflict to demonstrate resilience towards, and knowledge of, digital media literacy, this paper proposes the development of guidance for a “minimum basic technology infrastructure” which is supported by technology, regulation, and public awareness and education…(More)”.