My journey into medicine started when I lost my grandpa to cancer. I remember him dropping me off at my sixth-grade classroom in the morning, and by the time I got home, he was gone. I bought an oncology textbook not long after that. Science had always been my favorite subject, and I thought it could help me find some answers.

Years later, I was the doctor in the exam room, but I was still looking. The hospital where I worked was a national referral center for pediatric rare diseases, 95% of which have no cure. Every day, I was reminded how little medicine could explain about my patients’ conditions: the cellular dysfunction we could not see and the symptoms we could not explain.

Each of us has felt some version of that grief and frustration. People we love are misdiagnosed. We’re prescribed drugs that don’t work. It’s hard even to know how to talk about diseases like Alzheimer’s and Parkinson’s, where the science has been stuck for decades and hope seems so far from reach.

But I also believe all of this can change—not 50 years from now, but far sooner—if the scientific community works urgently and collectively to realize the promise of AI in human health.

We can already see glimmers of what’s possible. To take just one example, scientists have built frontier AI models that can generate entirely new kinds of proteins to target cancer cells and stop pathogens. The models work because they’ve been trained on huge volumes of data and developed a deep understanding of how proteins fold and function in the body. The same technology should be capable of modeling entire cells, tissues, organs, and potentially all of human biology.

Given that extraordinary promise—and the rapid progress we’re seeing right now—this is the moment for leaders across technology, scientific research, and philanthropy to lay the foundations for the next era of scientific discovery and cures. No one organization can do it alone. This is why we must bring the global community together to build an open data foundation for AI-accelerated biology. And it is why the institute I lead, Biohub, is announcing the Virtual Biology Initiative. 

Powerful cell models could fundamentally transform the process of discovery. For hundreds of years, scientific research has advanced by reducing questions to the simplest possible terms. We strip out confounding variables, remove complexity, and narrow the scope of our inquiry to processes that can be tested in a laboratory and understood within the length of a grant cycle. We’re left with knowledge that doesn’t represent our biology.

AI models aren’t subject to any of those constraints, which means they could finally give the scientific community a way to address the most difficult and urgent questions in human health. If AI can simulate and understand the immune system, it should be possible to engineer therapies to prevent diseases like cancer at the earliest stages. Or neurodegeneration. Or metabolic disorders. As far as we know, the possibilities for new cures would be limited only by the scale of the models.

But that also leads to the biggest challenge the field has yet to solve. Before AI can simulate biology, it needs to see biology, and the vast majority of cellular activity has never been observed or measured. Protein models typically trained on protein databases. Genomic models are generally trained on genomic databases. We still need an equivalent model for cells and the databases to train them—a massive, public resource that captures every type, behavior, and possible state they can occupy in the human body and other organisms.

To put it in motion, the scientific community will need to collaborate on an unprecedented scale.

Over the past decade, universities and research institutes all over the world have worked together to accelerate the scientific understanding of cellular biology, including its support of large-scale data generation projects such as the benchmark cell maps for humans and other organisms. We’ve also created repositories of cellular imaging data, and built one of the largest single-cell databases in the world. Last year, we brought public and private institutions together to launch the Billion Cells Project network, which is generating a massive open-source biological dataset.

The Virtual Biology Initiative will build on all of this work. To help jump-start a coordinated global effort, it begins with a $100 million commitment to fund data generation across the scientific community. Several other institutions are coming together with Biohub, including the Allen Institute, Arc Institute, Broad Institute, and Wellcome Sanger Institute—as well as consortia including the Human Cell Atlas and the Human Protein Atlas to coordinate a larger-scale effort. NVIDIA is partnering on this effort as well, and Renaissance Philanthropy will join to catalyze funding. 

Within Biohub, we’ll also continue developing frontier technologies to measure the cell. Imaging is a critical focus of this $400 million commitment: our roadmap includes microscopy to observe millions to billions of cells in living organisms, and cryo-electron tomography that can resolve atomic-level details in the cell. We also want to drive big advances in cell and tissue engineering, so researchers can run new kinds of experiments and measure biology we can’t access today.

If you have the resources to do or support biology research, I urge you to join in this effort. I am confident that AI models will solve mysteries in human health that the past century of research could not. We’ll reach those answers faster if we work together.

Open source technology is facilitating a new, more collaborative approach to research—one that brings multidisciplinary teams together to solve challenges that no institution can address alone. Experts have been talking about the promise of personalized medicine for decades. Through this effort and the AI models the data will power, I believe we can make that promise real for patients everywhere.

Whether they know it or not, millions of people—sick patients and worried spouses, families without answers, and people who haven’t begun to search—are counting on our success.