AI is smarter bioscience, but it hasn’t taken over | News

Joan Koerber-Walker, president and CEO of the Arizona Bioindustry Association (AZBio), has strong viewpoints on how artificial intelligence contributes to innovation in the biosciences.

“AI is an exciting area, but I do get frustrated when people say ‘AI is going to change the world,’ as if it’s some sort of magical solution,” she said. “Intelligence is not artificial. What we have today is pattern recognition and large data analysis, but AI isn’t thinking — yet. Tools like ChatGPT, for example, aren’t thinking; they’re regurgitating information.”

Turns out ChatGPT, when fed Koerber-Walker’s comments, has some opinions, as it were, about her, too.

“She’s correct in saying that AI does not ‘think’ in the human sense  —there’s no self-awareness or independent reasoning,” said the large language model. “This could be seen as ‘regurgitating,’ though I aim to offer contextually relevant and nuanced responses based on input.”

“However, I would also say that while AI might not think or reason like humans, it can still provide value in ways that mimic certain cognitive functions, such as pattern recognition, data synthesis and natural language processing,” ChatGPT continued. “These capabilities are transformative, especially in fields like healthcare, where AI can assist in analyzing large datasets, making connections and even offering diagnostic support — all of which would take humans much longer to do manually.”

Of course, they’re both right — human and android digs aside. In Tucson and Southern Arizona alone, AI is already playing a growing role in advancing bioscience research and innovation. Several projects and institutions in the region are utilizing AI to enhance medical research, diagnostics and healthcare delivery. 

But in each case, it’s still the humans in charge that are turning the resulting data into trailblazing bioscience advances.

For example, a team at the University of Arizona, led by Dr. Rui Chang, has been leveraging AI to accelerate the search for new treatments for Alzheimer’s disease. Using large datasets from more than 2,000 Alzheimer’s brain tissue samples, AI helps identify genetic pathways that could lead to drug development. 

But this AI-powered approach is groundbreaking primarily because it allows human researchers to analyze thousands of potential drug targets simultaneously, significantly speeding up the discovery process.

“The AI is a novel method that can tease apart massive data into a network model to provide a crystal-clear picture of upstream events, showing which upstream genes control important downstream genes,” Chang said in a press release last year. “With that model, we are able to pinpoint the upstream genes that trigger amyloid plaques and tau tangles downstream. These upstream genes may be better targets for potential therapies.”

Elsewhere, the Arizona Simulation Technology and Education Center (ASTEC) at the University of Arizona is developing an AI tool to train medical students in patient communication. Called the Artificially Intelligent Medical History Evaluation Instrument, or AIMHEI (pronounced “aim high”), the tool’s task is to provide real-time feedback during simulations to help students improve their interpersonal and diagnostic skills by analyzing their interactions with virtual patients​ — before turning them loose on the real ones.

“AI won’t replace doctors, but it will change the way we do our jobs,” said Dr. Allan Hamilton, ASTEC’s executive director. “Thirty percent of a doctor’s time is spent on record-keeping, which AI can handle. It can also help doctors catch things they might have missed. 

“For instance, AI can monitor a patient’s vitals in the ICU and detect subtle changes that might indicate trouble before a human would notice. AI enhances the efficiency and accuracy of care, but doctors are still making the final decisions. As I tell my residents, ‘Bots don’t go to jail; doctors do,’” he added, with a laugh.

Hamilton himself is a fascinating human. The Queens, New York-born physician initially graduated college as an English major, studying creative writing under Rod Serling, the famed creator of “The Twilight Zone.” That degree didn’t help him land a job, however, so he first found work as a janitor at a veterinary hospital. 

“While I was cleaning, I could see through the window and watch the surgeons doing their work,” he recalled. “One of the surgeons there took pity on me and said, ‘Look, if you get your chores done early, you can come in here and hold retractors, and I’ll explain what we’re doing.’ One thing led to another, and then I was on my way.”

“That surgeon gave me a chance, and it opened my eyes to medicine,” he continued. “From there, I went on to Harvard Medical School, and eventually I moved to Arizona, where I’ve spent a large part of my career.”

Along the way, Hamilton, now 72, also worked as a consultant for the TV medical drama “Grey’s Anatomy” for 20 seasons, a role that involved ensuring medical accuracy and providing insights into surgical procedures and medical practices portrayed on the show.

He said the operating room scenes we see on such shows, which today often depict surgeons interacting with AI or virtual and augmented reality (VR/AR) simulation tools to help guide them through delicate operations, actually reflect real-world changes happening in the biosciences.

“Yes, those things are really happening,” he confirmed. “For example, if I image your spine, I can then take that image to the operating room and ask AI to perfectly move a screw that I’m going to place, and to guide that drilling so that the placement of that screw is perfect in terms of the threads biting exactly the right density of bone.”

Not only can AI guide the surgeon to perform such a precise procedure, it can interact with other machinery to do much of the work itself. “There are small robots that AI can actually control to keep the drill moving in one set position as I advance it. And it’ll follow the most minute, tiniest little movements that we couldn’t really do with our hands. Right now it’s particularly used for placing of hardware, but I think you’ll see it start to be used for lots of other things, like removing brain tumors.”

For as much as AI tends to dominate the biosciences news these days, however, AZBio’s Koerber-Walker says that’s far from the only exciting thing happening in the current Tucson biotech field.

She also highlights the work of the Critical Path Institute (C-Path) headquartered in Tucson, which serves as a hub for collaboration in drug development. The institute brings together universities, researchers, companies and government bodies to share data and ideas in a pre-competitive space, aiming to avoid redundant research efforts and speed up the path to new treatments. 

“It’s the largest institute in the country where organizations can come together to share ideas,” said Koerber-Walker, noting that this collaboration is essential for advancing science more quickly and efficiently.

Koerber-Walker also cites the work of Dr. Jennifer Barton, executive director of the BIO5 Institute, who’s developing an innovative endoscope designed to diagnose ovarian cancers and cysts, a condition that is often difficult to detect early.

“It’s pretty cool,” she said. “Dr. Barton is working on a new medical device technology that could become the next standard of care. She’s combining optics with biomedical engineering to address issues that are very difficult to detect, like ovarian cancer, which is often hard to catch early.”

Another important player in Southern Arizona’s biotech landscape is Roche Tissue Diagnostics, based in Oro Valley. Koerber-Walker recounts how the company was founded by Dr. Tom Grogan, a UA pathologist, who sought a more efficient way to analyze cancer slides. His idea evolved into Ventana Medical Systems, which was later acquired by Roche for $3.6 billion. Roche continues to invest in Southern Arizona, making the region the global headquarters for its tissue diagnostics division.

“Tom Grogan is just a wonderful person, and he’s still living in Southern Arizona,” Koerber-Walker said. “He was a pathologist who saw an unmet need and created a technology that changed the way cancer diagnostics are done worldwide.”

As for AI and the expanded role it’s expected to play in the biosciences, Hamilton says he often wonders what his old mentor, Rod Serling, would make of robotics’ unbounded progress.

“He had a number of robotic questions that he posed in ‘The Twilight Zone,’” Hamilton said — like the episode where the head of a factory automates his entire business, replacing human workers with machines and another revolving around a scientist who builds a robot to replace him after his death. The robot adopts the personality of the scientist, showing how AI could simulate human behavior and interactions.

“As a surgeon, sometimes I think what would happen if I was developing all these groundbreaking new procedures that could only be done with AI,” he said. “It’s something you start to think about in the sixth or seventh chapter of your life. Could my work continue without me?”

Maybe AI will carry it on. Or maybe, more humanly, will the many students who’ve passed through the doctor’s class.