Resolving the Building Blocks of Matter with AI, Classical, and Quantum Computing 
Everything visible in the universe, from stars to living cells, is built from protons and neutrons, which are themselves made of quarks and gluons. Although the theory governing these fundamental constituents has been known for over 50 years, it remains unsolved: its equations are so complex that exact solutions are out of reach. This has driven physicists to develop powerful computational approaches, relying heavily on high-performance classical computers, a path that, while tremendously successful, is approaching fundamental limits. Quantum computing offers a new lens through which to study this theory, promising access to regimes that classical machines cannot reach. This project develops AI-powered tools to accelerate that discovery potential by building a unified pathway that harnesses the complementary strengths of both classical and quantum computing, bringing us closer to finally resolving some of the deepest open problems in fundamental physics.

Greener, Fairer Cities: Using Next-generation AI to Enable Resilient Urban Amenities
As climate change intensifies, cities face a "wicked paradox": while planting trees is essential to mitigate rising heat, these efforts are often hindered by tree mortality, management challenges, and the risk of resident displacement. This research project, led by Dr. Iryna Dronova and the Geospatial Innovation Facility, harnesses the power of Geospatial Foundation Models, a novel chapter in Artificial Intelligence, to analyze these complex urban challenges. Unlike traditional tools, this advanced AI will use decades of satellite data and geospatial datasets to help assess where green spaces are struggling and identify neighborhoods most at risk of heat and displacement. By studying a diverse set of U.S. cities, the team will develop data-driven strategies that balance ecological health with social justice considerations. Ultimately, this work aims to help inform holistic and resilient urban planning where nature and human well-being thrive together.

Using RNA Structure and AI to Design RNA Therapeutics that Treat Neurological Diseases
The power of AI and the success of AlphaFold have transformed the field of protein structure and engineering. Here, this project proposes to merge the power of AI with RNA structure data to design better RNA therapeutics. The team will use cell biology and biochemistry techniques to generate a large RNA structure database (>10,000 structures) then use emerging AI-driven optimization methods to design more effective and less toxic anti-sense oligonucleotides (ASOs) to treat neurological diseases. Because RNAs in the brain exist in a highly complex environment with many cell types, ASO design is computationally challenging. Thus, the project will robustly solve a large number of abundant RNA structures in different cell types, including in non-neuronal glial cells, and consideration of all cell types in the brain will limit toxic off-target effects. During the funding period, the team will generate preliminary data targeting neurological diseases, including Alzheimer’s disease, Parkinson’s disease, and brain cancer.

Deportation Data Project
The Deportation Data Project has been the only source of detailed Immigration and Customs Enforcement (ICE) data since 2025, releasing six data updates tracking every arrest, detention and deportation that ICE conducts. This work is not only the foundation for research in this area, but has also been the basis of over two thousand media reports. The datasets have helped uncover the basic ways that enforcement operates. For example, until very recently, most immigration “arrests” were not arrests at all. Instead, most immigration arrests were transfers to immigration detention of people already locked up in jails and prisons. Until the inauguration of the second Trump administration, ICE overwhelmingly targeted people already arrested by police on suspicion of a (usually minor) crime. This fact has changed as the new administration relies on indiscriminate raids—and this fact helps explain why those raids have been so novel and politically explosive.

From Behavioral Science Insight to Impact: Scaling a Digital Anti-Poverty Intervention for Millions
People across the world want to save money but consistently fall short but not because they lack the desire. Rather, the complexity of financial planning—the vast number of small, irregular expenses that are difficult to account for—create a "planning fallacy" or over-optimism about the future. The research team’s simple, low-cost "retrieval exercise" prompts people to think through their upcoming expenses category by category, to help households better meet their own savings goals. The results have been striking: in randomized evaluations in Zambia and Malawi, participants saved more, had more to consume during lean months, and had higher harvest yields. Now, the project goal is to broaden the reach of the research team’s paper-based tool to millions by adapting and delivering it digitally, powered by an AI-personalized interface. Seed funding will support a prototype and pilot in Rwanda, paving the way for global scale.

Low-Resistance Quantum Materials for Energy-Efficient Computing
As computing systems become more powerful and pervasive, an increasing fraction of their energy is spent moving information through the microscopic wiring that connects billions of devices inside a chip. As wire dimensions shrink, conventional metals such as copper become increasingly resistive, limiting speed and efficiency. This project investigates thin films of topological semimetals, an emerging class of quantum materials, as a new type of conductor that may carry electrical signals more efficiently at the nanoscale. By synthesizing these films under semiconductor-compatible conditions and relating their structure, interfaces, and composition to electrical transport, this project aims to establish design principles for nanoscale quantum conductors in future energy-efficient electronics. The project will also engage undergraduates in hands-on microelectronics research, providing training in quantum materials synthesis and nanofabrication.

Molecular Platforms for Nuclear Quantum Technologies
The Long research group at UC Berkeley is developing new molecular materials based on thorium to enable next-generation quantum technologies. Unlike conventional quantum systems that rely on electronic states, thorium contains a unique nuclear transition that is naturally protected from many sources of environmental noise, offering exceptional stability for quantum information storage and precision timekeeping. Using the tools of synthetic chemistry, researchers can precisely design the molecular environment surrounding thorium, enabling control over its optical and quantum properties. This project combines synthetic chemistry, spectroscopy, and quantum science to create molecular environments that can host and control these nuclear states. By developing new thorium-based materials and optical measurement techniques, the research aims to establish a foundation for future nuclear quantum memories and ultra-precise nuclear clocks.

Chemical Strategies for Targeting Protein Dark Matter Driving Human Disease
The development of AlphaFold, an AI-based structure prediction tool, marked a milestone for both drug discovery and basic science. In seconds, one can now predict the structure of many proteins with high accuracy – a critical first step in the design of new medicines. Intrinsically disordered proteins, however, remain an important exception – in general, these flexible biomolecules cannot be accurately predicted in most cases and are also often invisible in standard protein crystallography. This “blindness” represents a critical Achilles heel in modern- day drug discovery as there are no small molecules known to bind most of these regions with high affinity or selectivity. Critically, proteins with high structural disorder are also implicated in a variety of disease pathologies. Using advances in synthetic organic chemistry this project will develop new chemical tools and strategies to target these elusive structural motifs.

Recovering Footprints of Archaic Ancestry in Modern Genomes
Gene flow from archaic hominins, Neanderthals and Denisovans, into modern humans has shaped our genetic and phenotypic variation. DNA inherited from these groups has influenced a wide range of traits, including high-altitude adaptation, metabolism, and immune function. Yet many archaic populations that contributed to modern humans— particularly those lacking preserved fossils or DNA —remain uncharacterized. As a result, our picture of our past and the role of archaic ancestry in human evolution remains incomplete. This project will develop new computational approaches to recover and reconstruct these ancestral genetic lineages using DNA from people living today. These reconstructions will provide new insights into human origins, adaptation and disease. The project will also create open-source software and genomic resources that can be used by researchers worldwide. Together, this work will provide a powerful way to study human evolutionary history and understand how our past has influenced human health and diversity.

CRISPR Guided Insights into Microbial Methane Metabolism
Methane plays an important role in the global carbon cycle and is also an energy-rich, cheap, and abundant resource that can be used as a biofuel. The research team will use CRISPR-Cas genome editing tools to study the biology of microorganisms involved in the production, capture and conversion of methane. The work will enable deeper insights into enzymes and cofactors required for microbial methane metabolism, a process that is vital for the environment and the economy.

Human Reproduction in a Changing Climate
A major gap in scientific knowledge is what rising global temperatures will do to the biological ability of people to have children. Heat stress affects systems central to establishing and maintaining healthy pregnancies, and impairs fertility in animals. There is almost no credible causal evidence on how heat stress in humans affects key dimensions of reproduction: gamete development, ovulation, fertilization, implantation, and miscarriage. This research integrates clinical, administrative, and health-tracking data to prospectively follow several million adults attempting pregnancy and to study their fecundity and fertility. With rich climate characterization of their communities, we establish the causal effects of rising temperatures on a range of reproductive outcomes, from gamete quality through pregnancy survival. Demographic modeling will allow the team to establish how these effects cumulate—i.e., what are the implications of climate-caused changes in fecundity for the length of time people have to wait to become parents? For population fertility rates?

The Archaeology of Diet, Disease, and Migration at Nabataean Petra (Jordan)
This project investigates the early Nabataeans, an ancient Middle Eastern society who transported incense and other luxury goods from the Arabian Peninsula to the Mediterranean Sea. Among other achievements, the Nabataeans were the architects of the UNESCO World Heritage Site of Petra, located today in modern Jordan. This project draws on the archaeological sciences to investigate the early Nabataeans’ diet, biological health, and migration patterns using evidence excavated in Petra. This research will center the contributions of Berkeley students and early-career Jordanian and American scholars in archaeology, history, and the biological sciences. With the permission of Jordan’s Department of Antiquities, Berkeley will carry out this research in partnership with the American Center of Research, an overseas research center located in Jordan’s capital, Amman.

When Criminal Governance Ends: Economic Opportunity and Democratic Tradeoffs
This project examines what happens when the state dismantles criminal organizations that have long governed urban neighborhoods. El Salvador’s recent security crackdown sharply reduced gang territorial control, creating a rare opportunity to study the social, economic, and political consequences of ending criminal governance. The project asks whether formerly gang-controlled communities experience economic convergence, whether residents continue to face discrimination in labor markets because of where they live, and how exposure to criminal rule shapes citizens’ willingness to trade democratic protections for security. Using newly available census data, satellite imagery, cellphone mobility data, employer surveys, and public opinion surveys, the research will provide evidence on whether the end of gang control expands opportunity or instead gives rise to new forms of exclusion.

The Coral Rescuers
The Coral Rescuers is a 39-minute documentary and impact campaign to move students, divers, scientists, policymakers and communities from awareness to hands-on action for one of the world's most endangered ecosystems. This is not another doom documentary. It is a film about what is still possible, and about the people who refuse to give up. Shot across South Florida and the Keys, it follows scientists and reef guardians at the front line of a real-time crisis.

The Development of Reasoning about Religious Norms in Pluralistic Societies
This project will explore how children reason about norms that derive uniquely from a religion, including norms for how to dress, eat, and pray. Religious norms pose a fundamental challenge for peaceful co-existence, particularly within pluralistic societies, because the norms that one group sees as binding are routinely violated by others. Hindu youth in India, for example, must decide whether it is okay that Muslims—and even some Hindus—eat beef. Yet although disagreements over who religious norms apply to have fueled conflicts throughout history, there has been little psychological work on this issue. The team will interview youth from different religious and cultural backgrounds to explore how people develop answers to questions including to whom religious norms apply, why religious norms should be followed, and when—if ever—religious norms can be violated. The findings will provide insight into the psychological preconditions for religious tolerance and respect in pluralistic societies.

Computational Models of Language Acquisition in Embodied Interactions
Alane Suhr’s research focuses on studying collaborative, situated, language-based interactions between and with people. In these interactions, at least two different participants work together towards a shared goal, embodied in a shared environment; for example, consider two people designing and putting together a piece of furniture. Here, language is useful for coordinating participant plans and actions, and so it is crucial that participants are competent at using language with one another. Conversely, however, such embodied interactions also provide plentiful opportunities for learning language. This project studies, through a computational lens, what kinds of phenomena in and features of embodied interaction can serve as learning signals that a language learner can use to increase their agency through language use.

Engineering Precision Structures for Affordable Carbon Removal
Removing carbon dioxide directly from the air is one of the most promising tools for fighting climate change, but today’s capture systems are slow and energy-hungry. They rely on powders and pellets that take at least several minutes to soak up CO₂ and considerable heat to release it again. This project takes a different approach. Using a 3D-printing technique the research group invented and patented, the team rapidly prints intricate, sponge-like structures with an enormous internal surface area. Coating them with a thin layer of CO₂ sorbent material places every active site very close to flowing air, allowing the structures to capture CO₂ much faster than existing systems and with much less resistance to air-flow than conventional packed powder/pellet beds. The team will also test ways to release the captured CO₂ using light and gentle pressure changes instead of heat, sharply cutting energy use—a key step toward affordable, large-scale carbon removal.

Quantum Spectroscopy for Molecular Sensing and Control
Molecules are constantly vibrating, but whether strong light fields can synchronize those vibrations remains an open question. This project develops techniques in quantum spectroscopy to uncover how molecules shift from independent to collective vibrations in nanoscale optical cavities. By analyzing the quantum statistics of scattered light, we will determine whether molecules move independently or in concert, which is a critical distinction that shapes their optical response and chemical reactivity. This project will fabricate tunable plasmonic nanostructures tailored to these measurements and test whether tuning light-matter interactions can trigger predicted but unconfirmed collective behavior. Success would open new pathways for ultrasensitive molecular detection and light-driven control of chemical reactions, with applications in catalysis, energy conversion, and quantum technologies.

Building Fiscal Capacity in Fragile States
Many of the world's poorest countries struggle to collect taxes, leaving governments unable to fund basic public services. In a close collaboration with provincial tax authorities in Democratic Republic of the Congo, this project explores a new approach to property taxation using drone imagery and AI to identify and assess buildings using a simple digital system. A citywide randomized controlled experiment in Kananga, a city with a population of about 1.6 million, illustrated the viability of this system in low-income settings with weak state capacity. It also found that a progressive property tax system — one where wealthier property owners pay higher rates — raised 55% more revenue than a flat-rate system. The government scaled up the digital progressive tax system across Kananga in 2025 and 2026. Now, this project explores scale up to larger, more affluent cities such as Lubumbashi and Kolwezi, studying how the new system impacts property tax revenue as well as citizen demands for government spending and accountability.

How a Changing Planet Will Shape Future Pandemics
Climate change, deforestation, and the spread of human activity into wild areas are making it more likely that animal viruses will jump into people and spark the next pandemic. However, we have no reliable way to predict how these risks will shift in the coming decades, or where they will be greatest. This project will build computer models that forecast pandemic risk as the planet changes through 2100. By combining what we know about the environmental conditions that let different viruses jump to humans with projections of future climate, land use, and population, this project will map where and how often dangerous outbreaks are likely to begin. The team will then simulate how those outbreaks could spread worldwide to estimate the odds of a full pandemic. The result will be the first clear picture of how a changing world reshapes pandemic threats, helping direct where preparedness efforts should focus.

Establishing a Major Crop Model for the Epigenetic Regulation of Plant Regeneration
The Williams lab has discovered epigenetic mechanisms impacting the regeneration of plant tissues and organs using model lab species. Whether these mechanisms also impact regeneration in important crops remains unclear. The lab will study maize lines carrying changes in an epigenetic system that may influence how readily plants can regrow tissues and organs. Improving this process could help remove one of the biggest barriers to crop improvement, where promising gene editing advances often remain difficult to apply because many crop plants are hard to regenerate from edited cells. The project will generate key preliminary data on whether these maize lines show enhanced regeneration, as well as establishing a new crop model for the study of epigenetics and genome regulation.