This grant provides support to organize an annual, intensive, week-long training program at the University of California Berkeley introduces second- and third-year doctoral students interested in energy and environmental economics to state-of-the-art research in the field.  Each year approximately 60 doctoral students from universities across the country arrive in Berkeley to receive rigorous training in energy and environmental economics and have the opportunity to interact with leading faculty in the field. Nearly a semester’s worth of content is presented in a condensed period of time, helping students participants gain the knowledge necessary to undertake more in-depth research when they return to their home institutions. Grant funds will provide support for a number of student scholarships, with a specific focus on increasing the number of women and underrepresented minorities participants and strengthening mentorship activities over the course of the training program.

Individual states play a significant role in setting and implementing energy policies of all kinds, yet no comprehensive database exists of state-level energy policy initiatives and projects.  Without such a database, it is very difficult to understand how state policies harmonize, augment, or conflict with one another (or with federal policy), how state policies could be altered to better achieve overall energy goals, or how state policies differ (or do not) by region, geography, culture, industrial composition, and much more besides. This grants funds an initiative to address this information gap by creating an inventory that makes systematic information about state energy policies available to the research and policy communities. It will also engage scholars and decision-makers to use this new resource and begin to rigorously analyze the connections, overlaps, and relationships between different state energy policy regimes. After constructing the inventory, the team will analyze state policies across across four dimensions: emissions reduction targets, economic outcome targets, resilience planning, and cross-governmental coordination. Research findings will be published as reports and articles and disseminated at a final workshop and shared broadly, with a focus on engaging state and federal decision-makers.

The Journal of Open Source Software (JOSS) is a peer-reviewed academic journal that specializes in the publication of articles about open source scientific research software.  In JOSS, authors can submit a software project along with a description of its scientific use, capacities, limitations, the technical resources required to deploy it, how to access it, and its associated technical documentation.  Following submission, a ’software-native’ review (on GitHub) workflow enables community members to perform code review as well as review documentation and associated metadata. Once published, scientists who use, reuse, adapt, or modify a piece of JOSS-published software in their own research can then cite the JOSS Document Object Identifier (DOI), giving recognition to the software’s developers and creating corresponding professional incentives for scientists to contribute to the development of open source research software.  Funds from this grant support efforts to allow JOSS to better serve its authors and readers through improving and documenting a number of elements of the JOSS technical platform as well as generalizing the software review components for peer review of software outside of JOSS. Funded activities include planned improvements to the “bot” that automates much of the coordination and technical checking of software submitted to the journal, as well as the creation of developer guides, deployment recipes, and a reviewer management system.  The underlying JOSS infrastructure is itself an open source project, allowing other organizations interested in conducting automated software review to benefit from the JOSS team’s work.  

By combining archived software code with information on the operating system, application, drivers and other information about the computational environment in which a software program was originally run, The Emulation as a Service Infrastructure (EaaSI) can trick software into thinking it’s being run on the hardware for which it was built.  The result is a sort of software time machine, allowing historians and researchers to interact with decades old software just as users at the time interacted with it.  Even better, EaaSI’s emulations require no special equipment to execute.  Anyone with a web browser can connect to and use the service.  Funds from this grant, led and administered by Yale University Library (along with partner funding from the Andrew W. Mellon Foundation), support the continued expansion and development of the EaaSI ecosystem.  Planned activities include the introduction of new features, like the ability to model networked resources within emulated software and the emulation of mobile phone and tablet apps, as well efforts to grow the number of institutions hosting EaaSI nodes and to provide enhanced training and documentation for users.

Open source software (OSS) is an increasingly vital component of the scientific research enterprise, used in one form or another at every point in the research pipeline, from instrument calibration, to data collection and cleaning, to analysis and visualization, to archiving.  The centrality and importance of OSS has led to the realization within academic institutions of the need for formal mechanisms to identify and support those OSS projects most central to its researchers.  One model being explored is the creation of university Open Source Programs Offices (OSPO), special intra-university bodies charged with the support of important open source software.    This grant provides funding for the Open Source Contributor Fund at the Johns Hopkins University, a pilot initiative designed to enhance and deepen the university’s support for and engagement with faculty working on open source software projects.  Spearheaded by Associate Dean for Research Management G. Sayeed Choudhury out of the university’s new Open Source Programs Office, the Fund will make small grants of $10,000 to those open source software projects deemed to be most important to campus researchers.  The 16 projects supported over the grant period will be selected through a combination of voting and data analysis of research software dependencies. In addition to surfacing appropriate projects for support, the nomination and voting process will be used to canvass the use, development, and maintenance of open source software across Johns Hopkins. Choudhury and his team will also produce a playbook and other open source tools for use by other institutions who wish to implement similar programming in support of open source development.

This grant funds a project by computer scientist Jeannette Wing, Director of the Columbia University Data Science Institute and Professor of Computer Science, to adapt "formal methods” (the representation of computer science systems as mathematical objects) to AI systems.  Once the AI system, the input data, and the desired trust property are formally specified, the AI system can then be analyzed using mathematics, allowing a skilled analyst to rigorously prove or disprove statements about the system being represented. The technique holds obvious appeal for those concerned about the trustworthiness of AI systems, since a formal methods analysis has the potential to reveal how an AI system would or would not behave in novel situations.  Grant funds will support Wing’s attempts to extend formal methods theory to AI systems, including how to formally specify properties of AI systems like fairness, privacy, and robustness.  A particular focus of Wing’s work will be to better formally understand the relationships among such properties, in order to identify and generalize their commonalities and differences.  Wing will also work on trying to use formal methods to characterize, with respect to these trust properties, the relationship between AI systems and the datasets used for training and testing them.

Open and inexpensive hardware has the potential to revolutionize the creation and deployment of sensors and other scientific instruments, expanding access and lowering barriers to innovation in data-driven research methods.  Much of the activity within the open hardware movement has been on expanding the distributed production of hardware, through tools like the open licensing of hardware design and the creation of open 3-D      printing templates for instrument parts.  There has been comparatively less emphasis, however, on how to measure and ensure quality control in a distributed production process.  The widespread availability of inexpensive sensors will only revolutionize science, after all, if the sensors actually work.  This project by University of Washington researcher Nadya Peek will  improve our understanding of quality control in distributed manufacturing processes.  Over the course of the grant, Peek will engage in four streams of activity aimed at filling gaps in current open hardware calibration practices. First, she will develop a generalizable format for documenting the theoretical capabilities of a production machine like a consumer-grade 3D printer.  Second, once this format is created, Peek will use it to develop calibration software capable of verifying that a specific instance of that machine is performing to expectations and within acceptable error parameters.  Third, Peek will develop new software to monitor such machines in real time, ensuring that they are maintaining precision and calibration through the production process.  Fourth, she will develop low-barrier procedures for testing the precision and quality of the final output. In addition, Peek will also field a survey questioning how researchers in the open hardware community are adapting their distributed production processes in response to the shutdowns caused by the COVID-19 pandemic.  

The Internet Age is quickly giving way to the Age of the Algorithm.  Decision-makers of all kinds are increasingly turning to complex algorithmic methods to help them allocate resources, set policies, and assign risk.  Banks use algorithms to figure out how likely someone is to default on a loan. Online retailers use algorithms to decide which ads to display on your phone.  Pollsters use algorithms to determine who is and who is not likely to vote. Increasing reliance on algorithmic verdicts comes with risks of its own, however.  The worry is not so much that the algorithms might get things wrong—human judgement, after all, is hardly error free--but they might get things systematically wrong, disfavoring one group of people over another for arbitrary or irrelevant reasons.  The worry is that we might build algorithms, in other words, that are unfair. This grant funds efforts by a team led by Harvard computer scientist Cynthia Dwork that aim to address this issue. Dwork’s plans involve constructing new theoretical frameworks—based on rigorous mathematical notions called pseduorandomenss, latitude, and multicalibration--that can be used to define and evaluate whether an algorithm is fair or not.  Grant funds will allow Dwork to fully develop her theory, build some algorithms that meet that those characteristics described, and test them to see if they indeed perform as theory predicts.  If successful, the effort would constitute a significant stride forward in our understanding of an increasingly essential cog in the machinery of modern life. 

Virtual assistants like Amazon’s Alexa are quickly becoming the gateway to all other digital products and services. The convenience and power of these assistants has led more than 50 million American households to adopt a virtual assistant over the past two years, an astonishing pace. Yet the marketplace for virtual assistants is dominated by just two firms, with Amazon and Google controlling 95% of the market. Because virtual assistants usefully connect to other digitally enabled devices and services, and because they need to constantly listen for voice prompts from their owners, they are poised to collect unprecedented amounts of personal information about consumers, from listening in on all the Internet of Things devices in our houses, to our communications on social media, from email to Facebook, and from our search history and purchasing records to our finances and health. In addition, unlike browser-enabled searches that return a full page of search results, queries of a virtual assistant yield only one answer, giving them a unique ability to shape (and manipulate) what we encounter and what we know via the World Wide Web. Since virtual assistants are powerful intermediaries between consumers and the wider world, it would benefit all consumers if the market for these assistants was robust, giving consumers many options to choose from.       This grant funds a project by Monica Lam, professor of computer science and director of the Open Virtual Assistant Lab at Stanford University, to build and pilot the first prototype of an open source, privacy preserving virtual assistant. The project, if successful, promises to expand the options available to consumers and offer the ease and convenience of a first-class virtual assistant without the sacrifice of personal privacy or transparency.

Funds from this grant support the production and dissemination of a special series by Annual Reviews’s online publication Knowable Magazine that aims to provide fresh, science-based, and publicly accessible perspectives issues related to the COVID-19 pandemic. Called Reset: The Science of Crisis and Recovery, the 9-month series will feature articles, essays, profiles, interviews, infographics, video, and comics exploring the scholarly work that informs the best response to the coronavirus pandemic. Content will include reporting and expert commentary in the digital publication Knowable Magazine and republished content in diverse media outlets. In addition, the team at Knowable will launch a series of online events featuring renowned scholars from an array of fields discussing timely topics around COVID-19 and providing reliable, evidence-based information and guidance. Content will be informed by Annual Review’s roster of 51 journals and 1,000 scholars, scientists, and journalists. Additional grant funds will support efforts to reach new audiences, including potential content distribution partnerships with Yahoo! News, local radio stations, the Smithsonian, the Aspen Institute, and the Huffington Post, among others.