Human-Centered AI: Experts Weigh in on the Future of Healthcare and the Workforce

Blog  • 05/06/2022 Artificial Intelligence

By Michael Richards

Perhaps the new dictum of working with artificial intelligence should be: Do no harm. That’s according to testimony at the recent field hearing conducted by the U.S. Chamber AI Commission in Cleveland, OH. The Commission heard from panels of experts across civil society, government, academia, and industry on the impact of AI on healthcare and the workforce.

“When we graduate medical school, I did not take an oath to cure patients…The only oath that we took was to do no harm. Do you know why we take that oath? We can harm people,” said Dr. Serpil Erzurum, MD, Chief Research and Academic Officer at the Cleveland Clinic, in her testimony to the Commission. “We have in this room people who are going to have to make those policy decisions [on AI]…we have to hold true to that oath to ‘Do no harm’, because the patient is a vulnerable person.”

AI exists for the benefit of humans, assisting us with our tasks, improving our quality of life, and allowing us to reach new heights. In evaluating the benefits and risks of AI, we can’t forget that at the heart of artificial intelligence are human beings — that’s the message from experts in Cleveland, Ohio. “The value of an AI algorithm is not in its academic, scientific sophistication,” said Dr. Laura Jehi, MD, Chief Research Information Officer at the Cleveland Clinic. “Its value is in what it actually does to help people, and what it actually does at the end of the day to change healthcare for the better.”

“The scalability of the good that we can provide in people’s lives is imperative, not because it’s a business imperative, it is an ethical imperative,” Dr. Tom Mihaljevic, MD, CEO and President at the Cleveland Clinic, stated in his opening remarks.

According to the testimonies, AI has helped improve healthcare in innumerable ways, not only in helping patients and health providers, but also in reducing inequities, costs, and inefficiencies. “Without these algorithms, [nurses] had had to drop everything they were doing to focus on [COVID-19],” said Dr. Jehi. “The algorithms we created were the tool that allowed them to prioritize, allowed us to get to the sickest patients when they really needed us.”

On how AI afforded insights that enabled them to provide better and more equitable care, she added, “We saw how the rate of infection was growing at a much faster rate in our disadvantaged neighborhoods here around us…and because of it we ended up taking steps as a healthcare system to make our care more equitable in the region. We literally sent buses to go to those neighborhoods and offer testing for COVID. We put teams together to be available to make sure that those patients got connected back to our system.”

Carly Eckert, Executive Vice President at Olive, highlighted how AI can help address inefficiencies: “Administrative costs make up 1/3 of total healthcare expenditures…There’s an estimated $760 billion dollars a year wasted in inefficient or unnecessary services…There’s a real opportunity to use automation in the places where humans aren’t needed. This can boost productivity, decrease the number of mistakes humans make, and prevent costly errors. This also frees up our human workers to really do the tasks that we’re best at. It allows us to harness creativity to solve complex problems.”

At the same time, using AI in healthcare doesn’t mean removing the human factor. As Dr. Jehi testified, “There was a person with judgment that was looking what the AI algorithms were doing and making the final decision. But that human’s job was so much easier because of what the AI algorithms provided upfront, and that is the piece that cannot be lost in all of this.”

On the other hand, care providers also listed privacy, impersonality, and bias as potential drawbacks to AI-based healthcare systems. “Regulation, privacy, confidentiality are key, but they should not come at the expense of discovery,” Dr. Jehi stated.

On AI’s impact on the workforce, experts highlighted AI’s ability to create new opportunities, while cautioning its likely effect in displacing many jobs and the broader economy. As Ben Ko, General Manager at Kaleidoscope Innovation and Product Design, testified, “AI is often maligned as a job taker…How do we look at it not just as a job creator and also a job focuser? Humans doing what humans are best at.”

Related

“We are seeing some tremendous update in the workforce for this type of technology. Rather than a disruptor, we believe that this is actually an additive technology that will help the American workforce,” said Elizabeth Hyman, Chief Executive Officer of the XR Association.

Simultaneously,“We must be mindful of where the gain from these systems go,” cautioned Almutwakel Hassan, a student of statistics and machine learning at CMU. “The wealth gap in the country has been increasing and mass deployment of AI systems will almost certainly widen the gap between the most wealthy and the median American, if steps aren’t taken to prevent it.”

“We must ask ourselves if as a society we want AI automation to be framed as a way to replace people’s jobs to drive profit margins, or if we want AI automation to be framed as a way to improve working conditions and quality of life for people,” he stated.

To prepare the workforce for an AI-enabled economy, experts emphasized the need for education and training throughout industry and academia and in all age-groups. Cheryl Oldham, Vice President of Education Policy at the U.S. Chamber of Commerce and Senior Vice President at the U.S. Chamber Foundation, pleaded with the Commission not to forget about the importance of the K-12 system. “If we’re trying to address our workforce challenges by starting post-high school, we’re failing.”

“Most AI curricula that are out there are relegated to after-school programs and summer camps and schools that have the resources to actually deliver technology education through the lens of computer science,” testified Alex Kotran, CEO of the AI Education Project. “The challenge of that is that computer science is an imperfect delivery vehicle for learning about AI, because many schools don’t have computer science teachers whatsoever. And even schools that do have computer science teachers lack equitable access to those programs.”

“If we want to be a technology-proficient workforce, if we want our children to be prepared to compete in a 21st century advanced manufacturing economy and equipped with the skill sets for jobs that have yet to even be invented, we have to do a better job in Ohio and through the U.S. to integrate computer science offering at the K-12 level,” Rick Carfagna, Vice President of Government Affairs at the Ohio Chamber of Commerce, said in his remarks.

Oldham added, “We need to do two things – we need to minimize any negative disruption and then we need to put AI to work for the American workforce…We need to proactively lean into workforce development and make it more authentically employer-led, if we’re going to minimize any labor market disruptions while also building new and effective pathways that will lead to AI-related jobs.”

Closing out the testimonies, Hassan advised the Commission to broaden its reach, “I specifically ask that you reach out to university students as well both at the undergraduate and graduate level, because these students are the future leaders of the field, and we’ll be the ones researching innovations for the technology in the future.”

To continue exploring critical issues around AI, the U.S. Chamber AI Commission will host further field hearings in the U.S. and abroad to hear from experts on a range of topics. The next hearing will be held in Palo Alto, CA, examining the benefits of AI, America’s global competitiveness, the future of work, and concerns regarding bias. The final hearing will take place in London on June 13.

Learn more about the AI Commission here.