It is an exciting area of academic study – one that has great potential for big impact. But the way modern technology has been designed has resulted in a myriad of unanticipated and unwanted outcomes.
In the Visual Digital Futures profile we seek to develop the next generation of world-changing technologies from a different perspective, one that asks the question: how do we design technologies without endangering our environment, our democracies, our values, and our future?
To address this question we are bringing together researchers and scholars from across LiU who study, design, and build digital technologies. This nexus of diverse expertise allows us to not only imagine new, powerful tools and algorithms, but to thoughtfully and purposefully design them within the pressing constraints of our society and our world. We seek to imagine new and different futures we might want, and to work to translate those futures into actionable, engineering efforts. We emphasize the importance of human decision-making by centering people through visual and interactive tools.
Profile area - Visual Digital Futures
Visual digital futures is a world-leading research initiative to fundamentally change how we design technology for society’s most important challenges.
Visual digital futures overview
Technology has fundamentally changed our lives. It has altered how we make scientific discoveries, how we conduct our work, how we take care of ourselves, and how we socialize and connect.
Our goal
To design technologies that give people the capabilities, agencies, and responsibilities for tackling pressing societal needs in order to shape the futures we want.
A1: Decision Support for a Sustainable Future
The unprecedented impacts of climate change and biodiversity loss require radically new approaches and tools for making decisions and enacting necessary change. We contend that executing effective climate change mitigation and adaptation strategies necessitates next-generation decision-support frameworks for complex, data-intense scenarios that render the causes and impacts of climate change legible, explorable, usable, and transparent for scientists and decision makers. Developing these frameworks brings together competencies from climate science and policy research, biodiversity and ecology, environmental change, visualization, computer vision, and AI. At the intersection of these competencies are basic research questions: How can machine learning and AI be transparently integrated within interactive visualization systems? What kinds of visualization can support informed decision making? To what extent can visual analysis tools support data analysis and decision making related to both local and global challenges?
A2: Visualization for STEM Education
Interpreting and interacting with visual information to construct scientific, technological, engineering, and mathematical (STEM) knowledge underpins our understanding of the world. The importance of STEM in Sweden is emerging as crucial for informing the future educational infrastructure, developing workforce, and for eliciting an informed citizenry. Visualization is fundamental for engaging in STEM, and as a means to catalyze curiosity, provide learning opportunities, solve problems and make informed decisions in a digital world. For example, understanding the climate crisis, interacting with AI to scaffold knowledge construction, and designing therapeutic medicinal interventions all involve interconnecting visualization and STEM for engagement, education and action. We have, however, only begun to rise to the challenge of solving how novel visual digital environments can be leveraged for meaningful and sustained STEM engagement and education. This challenge requires unique approaches that focus on how the power of visualization can be harnessed to evoke personal relevance, spawn curiosity, integrate intergenerational learning, as well as incorporate evolving visual literacy skills for strengthening STEM engagement and education.
A3: Critical Visualization
For many years, the visual has been used as a placeholder for Truth, imbuing modern data visualizations and digital images with authority and reverence. But as society increasingly turns to data, algorithms, and visualizations to support decision making, we are also becoming aware of the ways that they are partial and imperfect, reinforcing existing biases, hegemonic viewpoints, and blind-spots. New perspectives on how we come to know and record things about the world, however, are bringing forward opportunities for different people, communities, and considerations to shape future technologies. What is needed is a critical dialogue around the ethical and societal consequences of our reliance on visual technologies. This dialogue – between humanists, technologists, and scientists – might ask: How can we develop a visual literacy that supports the public in critically reading the visual representations of the world that they meet, from election results, to climate change, to medical information and more? How do we develop tools that make clear that visualizations are partial and situated, and reveal which perspectives might be missing? How can we rethink the ways that visual systems are designed to ensure that future technologies are inclusive, responsible, and fair?
A4: Building Trust in AI-Powered Precision Health
AI has transformed healthcare into a more personalized and precise field by providing capabilities to sift through vast and complex datasets. Now, generative AI is set to revolutionize the field by creating synthetic medical data for research, aiding drug discovery through the simulation of molecular structures, and advancing the precision of healthcare utilizing next generation medical imaging and multi-omics data. While the goal of this revolution is to empower healthcare professionals, the overwhelming volume of data and the inherent opaqueness of the algorithms makes understanding and trusting AI tools in clinical settings a significant challenge for adoption. Overcoming this requires research into not only how we design more intuitive interfaces and transparent AI algorithms, but also into understanding and acknowledging socio-technical concerns such as the limits of data and algorithms for clinical settings. Collaborations between technologists, medical researchers, and sociologists are necessary to ensure that next-generation AI tools will improve healthcare outcomes, optimize resource allocation, and reduce costs for everyone in safe, effective, and trustworthy ways.
A5: Intelligent and Responsible Materials Discovery
The discovery of novel materials that enable faster, smaller, and more energy-efficient devices is vital for meeting the needs of a broad range of critical industries, from sustainable computing and energy storage, to environmental sensing and the monitoring of greenhouse gasses. Traditional approaches of discovering new materials through brute-force trial-and-error experiments, however, are too slow, while straight-forward approaches based on simulations are computationally prohibitive with too many candidates to sift through using heuristics. What is needed is nothing short of an AI revolution: combining the predictive power of data-driven ML models with modeling knowledge from theoretical physics and material science, integrated into visual analysis tools that can aid scientists in the identification of key patterns, structures, and relationships within materials data. AI-driven discovery tools have the potential to facilitate more informed decision-making about good candidate materials, to speed up the development of cutting-edge semiconductors and catalysts, and to ultimately allow for the accelerated knowledge-based design of novel materials that decrease resource use, both in terms of energy and critical raw materials. How we design and build these tools, and do so in ways that responsibly consider the societal trade-offs of the massive energy consumption of AI algorithms, is an open question.
