What is Funding the Frontier?
Funding the Frontier (FtF) is a new machine learning-based tool designed to illuminate how research funding translates into real-world outcomes. While traditional assessments focus on grants and scholarly papers, FtF traces a broader impact—from patents and clinical trials to policies and news coverage—across a vast network of data. The goal is to provide funders, policymakers, universities, and researchers with a multidimensional view of how investment shapes science and society.
How FtF works: linking investments to innovations
FtF builds on one of the world’s largest datasets, drawing from platforms such as Dimensions, SciSciNet, and Altmetric. In total, the system connects over 7 million research grants to 140 million papers, 160 million patents, 10.9 million policy documents, 800,000 clinical trials, and 5.8 million news articles, weaving these items into 1.8 billion citation relationships. This expansive mapping enables users to see not just which projects generate publications, but which drive tangible innovations, inform policy, or influence public understanding.
Applications for decision-makers
According to the study’s first author, Yifang Wang of Florida State University, FtF could help decision-makers adopt a more comprehensive view of impact. By visualising funding distribution across fields, institutions, gender, and career stage, FtF highlights who receives funding and where inequities might exist. Such insights could guide funding strategies toward higher-impact pathways and improve forecasting for future opportunities.
What experts are saying
Independent researchers view FtF as a promising advance in scientometrics. James Wilsdon of University College London notes that FtF’s breadth—combining diverse data elements in a single framework—distinguishes it from many prior efforts. Vincent Traag of Leiden University praises the user interface for enabling exploration of outputs, outcomes, and impacts in an intuitive way.
Potential concerns and cautions
Not all experts agree that more data and predictive models should steer funding decisions. Critics worry about overreliance on metrics, which could bias toward “safe” projects and undervalue curiosity-driven or long-term science. There is concern that a self-fulfilling prophecy could arise if future investments are guided primarily by past performance rather than transformative potential. Ethicists caution that using past impact as a proxy for future success might entrench the status quo and overlook disruptive research that takes longer to materialize.
Another challenge is uncertainty. AI-based predictions are inherently partial and depend on how data is collected and interpreted. Some argue that FtF should serve as a supplementary tool rather than a decisive arbiter of funding decisions, providing context while leaving room for judgement and exploration beyond historical patterns.
Balancing innovation and accountability
Proponents argue that FtF offers a practical path to more transparent funding decisions. By linking grants to a spectrum of outcomes—from publications to health impacts—funders can better assess where investments yield broad societal benefits. Critics, meanwhile, stress the need for safeguards: continuous evaluation of the model’s assumptions, validation against real-world outcomes, and measures to avoid narrowing the research agenda to what is readily measurable.
What’s next for FtF?
As FtF moves toward broader adoption, stakeholders will likely push for ongoing refinement, including richer contextualization of grant proposals, better handling of causal inference, and clearer communication about uncertainty. The tool’s developers emphasize that FtF should be used as a supportive, not determinative, element in shaping research strategies—an aid for more informed, nuanced conversations about the future of science funding.
