A little over a year ago, the 37-year-old from Renton, Washington, was barely conscious, ravaged by a rare blood disorder called POEMS syndrome. The disease had left him with numb limbs, an enlarged heart, and failing kidneys. Doctors drained liters of fluid from his abdomen every few days. He was too weak for a stem cell transplant—his last hope for remission.
“I gave up,” Coates said. “I just thought the end was inevitable.”
But his girlfriend, Tara Theobald, refused to accept defeat. She sent a desperate email to Dr. David Fajgenbaum, a physician-researcher in Philadelphia whom they had met at a rare disease summit a year earlier.
By the next morning, Dr. Fajgenbaum replied with an unconventional treatment plan: a mix of chemotherapy, immunotherapy, and steroids—none of which had been tried before for Coates’ condition.
Within a week, Coates began responding. Four months later, he was strong enough for the stem cell transplant. Today, he’s in remission.
The lifesaving idea didn’t come from a doctor’s intuition. It came from artificial intelligence.
See also: The Bizarre Meaning Behind the Blue Spiral in the Sky—It’s Disturbing the Internet
A.I.’s Hidden Cures
Around the world, scientists are using machine learning to scan existing medications for potential new uses—a process called drug repurposing. While not new, A.I. is accelerating discoveries, offering hope for patients with rare and deadly diseases.
Dr. Fajgenbaum, a University of Pennsylvania professor and a Castleman disease survivor, leads one such effort. His nonprofit, Every Cure, uses A.I. to analyze thousands of drugs against 18,500 diseases, ranking them by likelihood of effectiveness.
“I had this really clear realization that I didn’t have a billion dollars and 10 years to create some new drug from scratch,” he said.
His own life was saved by sirolimus, a generic drug typically used to prevent organ rejection. Now, his team’s A.I. platform has identified treatments for aggressive cancers, inflammatory disorders, and neurological conditions—often with stunning success.
See also: Snow White’s Real Story: Folklore, Myth, and the Power of Fairytales
‘The Side Effect You’re Looking For’
In Birmingham, Alabama, another A.I. model developed by Dr. Matt Might at the University of Alabama suggested an unusual fix for a 19-year-old suffering from chronic vomiting: inhaling isopropyl alcohol.
“It worked instantly,” Dr. Might said.
His team has since found other surprising matches:
- Amphetamines (typically for ADHD) relieved periodic paralysis in children with a rare genetic disorder.
- A Parkinson’s drug improved speech and movement in patients with a neurological condition.
- Guanfacine, a blood pressure medication, restored mobility in a pediatric patient.
“Many drugs do more than one thing,” Dr. Might explained. “If you comb through enough drugs, you eventually find the side effect you’re looking for—and then that becomes the main effect.”
See also: NASA Astronauts’ Stunning Transformation After 9 Months in Space – Before and After Photos
Why Big Pharma Isn’t Interested
Despite its promise, drug repurposing faces a major hurdle: profit.
“If you use A.I. to come up with a new drug, you can make lots of money. If you use A.I. to find a new use for an old, inexpensive drug, no one makes any money off of it,” Dr. Fajgenbaum said.
To fund research, Every Cure secured $100 million from TED’s Audacious Project and the Advanced Research Projects Agency for Health (ARPA-H). The goal? Clinical trials for repurposed drugs.
See also: When Choice Ends: How Abortion Bans Impact Black, Hispanic, and Low-Income Women
‘Someone Had to Be the First to Try’
Not every A.I.-suggested treatment works—and doctors remain crucial in vetting risks.
When Dr. Fajgenbaum’s model proposed adalimumab (an arthritis drug) for a Castleman patient, Dr. Luke Chen was skeptical. “I didn’t think it was going to work,” he admitted.
But with no options left, they tried it. Weeks later, the patient was in remission—a case now published in The New England Journal of Medicine.
Similarly, when Dr. Wayne Gao in Washington first saw the aggressive drug combination suggested for Joseph Coates, he thought it was “a little bit crazy.” But with Coates facing death, “someone had to be the first to try.”
See also: Tesla Can’t Keep It Together—Literally, and Recalls Thousands of Cybertrucks
A Second Chance at Life
Last month, Coates visited Dr. Fajgenbaum in Philadelphia—healthy, strong, and grateful.
“I feel just fine,” he said, smiling.
For researchers, his recovery raises a tantalizing question: How many more cures are hiding in plain sight?
“This is one example of A.I. that we don’t have to fear,” said Dr. Grant Mitchell, an Every Cure co-founder. “This one’s going to help a lot of people.”