A team of Weill Cornell Medicine investigators is working to cross-train the next generation of cancer researchers in cancer biology and the use of artificial intelligence tools for research.
"AI is transforming our world—how we work, live and conduct research," said Dr. Olivier Elemento , director of the Englander Institute for Precision Medicine and a professor of systems and computational biomedicine. He added that "cancer is uniquely positioned to benefit, because we now have massive datasets spanning genomics, imaging and clinical outcomes that AI can finally put to use."
In an editorial published April 13 in Cancer Discovery, a journal of the American Association for Cancer Research, Dr. Elemento and Dr. Paraskevi Giannakakou , a professor of pharmacology in medicine and member of the Sandra and Edward Meyer Cancer Center at Weill Cornell Medicine, lay out a vision for a new generation of researchers who are "bilingual" in AI tools and cancer biology. They describe a dual-track program that would train clinical oncology fellows or cancer research scientists to use AI large language models and train computational scientists in cancer biology. The goal is to develop a generation of scientists who can use the latest tools to deliver on the promise of personalized cancer therapies.
"The moment a patient gets a cancer diagnosis, we would like to molecularly characterize the tumor, use all the knowledge that is out there—which is what AI large language models do—and see what therapies might be available and how we can customize treatment," Dr. Giannakakou said.
Oncology is already generating vast amounts of molecular and genetic data from patient tumors to personalize care, Dr. Giannakakou said. Now, she and Dr. Elemento are hoping to supercharge that capability by building a workforce that can leverage AI tools. The team is already building the dual training program and pursuing grants to fund it. The program would assign each trainee to a pair of mentors—a computational mentor and a cancer biologist or clinical oncologist mentor.
But powerful tools demand careful hands. Dr. Elemento, who describes himself as a "super user" of AI for research, said part of the training will be teaching the trainees how to properly supervise AI tools to avoid ethical breaches, protect patient privacy, and detect inaccurate results. He noted there have been examples of fake papers with synthetically generated data.
"It's important that people learn how to ethically use these extremely powerful tools," said Dr. Elemento, who is also a member of the Meyer Cancer Center at Weill Cornell Medicine.
Weill Cornell Medicine is well-positioned to lead this charge, according to the team. The Englander Institute of Precision Medicine is already offering "AI clinics" where AI-fluent investigators train their colleagues in person and via Zoom. An upcoming workshop will focus on securely using AI to analyze information from medical records. Meanwhile, Weill Cornell's AI to Advance Medicine initiative aims to provide the institutional infrastructure and services needed to support safe, effective AI adoption across faculty, staff and students.
Weill Cornell Medicine's robust expertise in cancer biology and clinical medicine will also ensure that AI is answering the most important questions for patients with cancer, Dr. Giannakakou said.
The team emphasized the urgency of training the next generation of researchers to use AI. Dr. Giannakakou noted that pharmaceutical companies are already deploying AI tools and agents to design clinical trials, monitor drug safety and help meet regulatory requirements, and that those working in the field need to be prepared to work with these tools.
"We need federal, private and institutional foundation investments in training programs to create AI-cancer biology bilingual scientists like we are creating at Weill Cornell Medicine," Dr. Giannakakou said. "We do not want to leave a generation of scientists behind."
