Harnessing the power of players to solve the mysteries of cancer

Harnessing the power of players to solve the mysteries of cancer

Cancer is malignant. In 2025, it is expected to cause more than 618,000 U.S. deaths — nearly twice the combined populations of Merced and Modesto. Each year, nearly half of this nation, young and old, is touched by the disease through a personal diagnosis or an affected loved one.

Jeff Yoshimi joined the 50% when his wife Sandy learned she had breast cancer. The fused cells had spread to some lymph nodes.

Beside Sandy during one of many overnight hospital stays, Yoshimi went in and out of sleep, reflecting on ideas drawn from his profession and his passions: a Ph.D. In philosophy, you study how people perceive and experience the world, and the expertise involved in clarifying difficult concepts by visualizing them in game-like ways.

How could he, a UC Merced cognitive science professor, fight a terrible disease?

Suddenly the pieces snapped together in his head. Yoshimi grabbed his laptop, walked across the street to a late-night cafe and started typing. He knew cancer could be curable, but it takes so many forms and changes so often that it is incredibly difficult to fight off. But what if the brain power available for cancer research has been increased exponentially? What if that power came from millions of minds who love unraveling mysteries, making buttons, and wiping out villains?

What if Yoshimi thought we'd expanded the cancer battle with a series of really great video games? From that first flash in 2013 and over 11 years, he expanded the idea into a compelling action plan and put it into a book. “Gaming Cancer,” published by MIT Press. In it, he claims that gamified citizen science, partnered with the rising potential of artificial intelligence, can help drive significant advances in cancer research.

Sandy Yoshimi is alive and well today after enduring a grueling journey of surgery, chemo, radiation and pills. However, cancer did not end with the family. Sandy Yoshimi's sister died not long after Jeff Yoshimi wrote the book. In 2023, bile duct cancer took over Sandy's father's life.

“Much of this book was written in cancer wards and chemo rooms,” Yoshimi says in the first chapter.

The book's subtitle describes the goal: "How creating and playing video games can accelerate scientific discovery." With a breezy writing style, Yoshimi guides readers not just on how it can be done, but why it can be doneshouldbe done.

Cancer-related problems are messy and complex, but in some cases they have a core logic that can be presented in a game context with its simplified rules. “

Jeff Yoshimi

Citizen science games are out there, although on a much smaller scale than Yoshimi in the book. An example: “Foldit” makes a video game out of restructuring proteins. Results from this “protein folding” game can help scientists understand cellular mutation and runaway replication of diseased tissue. It can help perfect therapeutic drugs.

In 2020, around 750,000 users played “Foldit”. An impressive number, but Yoshimi is thinking bigger. Like “Baldur’s Gate” bigger. “Candy Crush” bigger.

Yoshimi has practical experience with game style programs. In 2008, he created Simbrain, where users design neural networks with simplified visuals. He uses Simbrain in his courses and research, helping students and colleagues grasp aspects of neuroscience, psychology and artificial intelligence.

In an interview, Yoshimi, a UC Merced founding faculty member who came on board in 2004, spoke about the knowledge that could be unleashed by scaling citizen science. He told a story about going to a restaurant with his father and seeing a boy whose knee hit like a jackhammer.

"My dad said, 'I wonder if we could make a little generator, attach it to this knee and hold all this energy," Yoshimi said.

"That stuck with me. Humans have these exquisitely evolved brains and we enjoy solving puzzles. It's good for the species. So we have this huge reservoir of untapped power-solving power, this Earth-wide computational effort that's being used on Grand Theft Auto."

“What if we diverted some of them to problems of wider concern, like cancer?”

In the book, Yoshimi discusses the “meta-game” – the process of creating this suite of powerful, compelling games. He describes everything they need: crack programmers, visionary leaders and deep financiers, as well as reward structures for users and smooth connections between players and labs. At the center of it all, Yoshimi imagines Simbody, a game engine that simulates biological systems from the entire body down to the nanoscale.

Start with a cancer reconnaissance. Break down the problem until there is a task that a game could address. Plug it in. Yoshimi said the suite, which he called "Cancer Wars," could use any game style: action, adventure, role-playing, first-person shooter, strategy, sports.

Just a few years ago, artificial intelligence, which Yoshimi has been studying for years, hit the turbos with the arrival of large language models like Chatgpt. Suddenly it seemed that AI was going to change the game. Why turn to people for crowdsourcing when LLMs have trillions of information crammed into their silicon brains?

Why not do both?

According to Yoshimi, many of the existing citizen science games have AI elements that work hand-in-hand with the player, matching the data processing power of the former with the intuition power of the latter.

“We want to have human symbiosis,” he said. "AI can do raw numbers and statistical generalization. Humans see the bigger picture, the relevance of one thing to another, the creative insight that a machine has a harder time capturing."

Remember Capt. Picard and data, Yoshimi said. Luke Skywalker and C3po. Kirk and Spock.

“AI has made great leaps, but it's not at a point where we press a button and Chatgpt thinks about it for a month and solves the problem,” he said. “The best games incorporate AI with human intelligence as much as possible, so the experience is seamless.”

Yoshimi hopes “Gaming Cancer” provides a roadmap for scaling up video games in the fight against cancer. He added that eradicating the disease is a worthy goal, but certainly not the only one. There are countless important victories within reach. Detection and treatments for hundreds of cancer types require attention. Games can promote an understanding of clinical research and encourage lifestyle choices that can reduce the risk of cancer.

Because this is cancer we are talking about. Only heart disease kills more Americans. Why not bring a legion of video gamers into the fight?

“It’s worth the effort,” Yoshimi said. “Even if you don’t hit the moonshot, all the intermediate shots are valuable.”

Sources: