-Ruby Barcklay and Linda Vu
Artificial intelligence is everywhere we look — featured in headlines, scientific publications, scientific conferences and workshops, and increasingly in proposal calls. While AI might appear to be a recent phenomenon, its core technologies and methods were built over decades of scientific exploration. At Berkeley Lab, foundational advances in computational science, mathematics, and data analysis paved the way for modern AI, shaping the tools and approaches now powering data-driven discovery. Today, researchers across the Lab are continuing this tradition — accelerating AI adoption in science, expanding collaborative efforts, updating research strategies, and leading key national projects to support the Department of Energy’s Genesis Mission. Through this active engagement, Berkeley Lab is laying the groundwork for the next generation of AI-enabled scientific breakthroughs.
Building on our AI Past
The mathematical and computational foundations of artificial intelligence were established long before the term was widely known. At Berkeley Lab, researchers began applying mathematical and computational techniques as early as the 1930s — simulating particle paths to improve cancer therapies, developing statistical methods for complex experiments, and building systems to process massive datasets. Notably, they became leaders in applying and refining approaches such as Monte Carlo modeling, which uses thousands of simulated trials to explore possible outcomes and guided research in accelerator physics, nuclear instrumentation, and large-scale data analysis. These efforts helped lay the groundwork for scientific computing traditions that would later inform developments in artificial intelligence and machine learning.
As scientific challenges evolved, so did Berkeley Lab’s leadership in advancing computational methods. The Lab’s world-class scientific facilities and data resources — including supercomputers at NERSC and real-time data streaming via ESnet — enabled new approaches to data-intensive research. Together, these assets helped establish integrated platforms that link instrumentation, computing, and analytical workflows across disciplines.
Strategic investments over the decades strengthened the Lab’s capabilities in data science, advanced computing, and interdisciplinary collaboration. Notable advances included applying machine learning techniques for breakthroughs in materials science, biology, and energy; launching virtual deep learning schools to train scientists in AI methods; and convening researchers and industry leaders through gatherings like the Monterey Data Conference to drive innovation and knowledge sharing.
The Lab’s expertise in managing unique, high-quality scientific datasets — from the Advanced Light Source to the DOE Joint Genome Institute and the Materials Project — provided the critical data foundation for domain-specific and physics-informed AI. Models and analytical methods built on this foundation powered automation, predictive simulations, and control systems throughout user facilities.
Berkeley Lab’s historic achievements in computational science, data analysis, and automation chart a legacy of innovation—a legacy that continues to inform and shape scientific approaches to discovery and data-driven research.
Engaging in our AI Present
Today, AI is integral to Berkeley Lab, and its adoption is accelerating. Researchers are incorporating AI into their work. Research Areas and user facilities have updated how AI fits into their strategies. The Office of the Deputy Director for Research has convened and facilitated many AI meetings to encourage collaboration and the sharing of best practices. IT has jumped to meet the challenge, offering AI tools, skill building, and services to researchers across the Lab.
Now, the Lab’s AI capabilities have been called into service with the Department of Energy’s Genesis Mission, a national initiative to advance artificial intelligence and accelerate science, energy, and national security solutions. Berkeley Lab is leading or playing a key role in three of the Genesis Mission’s initial projects, sometimes referred to as “seed projects”:
- Multi-Office particle Accelerator Team (MOAT): MOAT will deploy AI tools to optimize how accelerators are designed and run, making them more powerful and efficient, speeding progress on national priorities. The project will take advantage of troves of experimental data, simulations, and expertise from across the DOE Office of Science’s accelerators and light sources.
- Synergistic Neutron and Photon Autonomous Science – Imaging (SYNAPS-I): This multi-lab initiative accelerates breakthroughs at leading X-ray and neutron user facilities in the U.S. by making discoveries from unprecedented volumes of data. Working directly with industry partners, SYNAPS-I will transform petabytes of imaging data into actionable knowledge that can be applied to critical technologies in microelectronics, medicine, advanced manufacturing, and energy security.
- Orchestrated Platform for Autonomous Laboratories to Accelerate AI-Driven BioDesign (OPAL): The OPAL project is using robotic systems, AI agents and models, and standardized data-sharing platforms to accelerate the biotechnology pipeline all the way from gene discovery to commercialized technology. The OPAL team’s Genesis project will develop powerful, general-purpose biology AI models that can be tailored for specific applications with additional training and eventually control AI agents to manage investigations autonomously. The model will help unleash the full potential of biology for manufacturing fuels, chemicals, consumer goods, agriculture, and critical mineral recovery.
In addition, Berkeley Lab is contributing to other projects focused on AI code development, critical minerals and materials, cosmology, microelectronics, and quantum algorithms.
Carol Burns, the Lab’s Deputy Director for Research, said, “With our scientific user facilities, large-scale experiments, supercomputers, expertise in science, math, and computation, AI-ready data, and partnerships, we are in a good position to take advantage of these opportunities.”
Preparing for our AI Future
Branden Brough, senior advisor for strategy with the Lab Directorate, who has been coordinating the Lab’s AI efforts, notes that these Genesis projects are just a first step, and that the focus on AI will continue to ramp up. It is a top priority for the DOE; the Administration’s FY27 R&D priorities memo in late September features AI prominently. It is seen as a “force multiplier” that can help achieve goals faster, better, and more efficiently.
“We need to be thinking about how we build upon this foundation,” said Branden.
He also notes that speed is of the essence. Said Branden, “The DOE’s AI aspirations are ambitious, with expectations that AI will be transformative across the scientific ecosystem. With those ambitions comes urgency. Waiting for a conference six months from now, or for the next proposal call, will be too long.” He notes that some of the initial Genesis projects have first deliverables due in a few months, with major deliverables due in the summer.
He continued, “I’d encourage Berkeley Lab researchers to find ways to make progress now. In almost every domain, we have an opportunity to make an impact through the integration of science and AI. There are a lot of things happening and we will try to use the Lab’s communication channels to ensure everyone knows what is going on but people should proactively take actions to make sure they’re in the loop. Start initiating team building with internal and external partners across domains. Experiment with new ways of using data and models in your work. When the proposal calls come out, we want to be able to say we have a team, we’ve made progress, and we can show results quickly.”
