As artificial intelligence and generative AI such as ChatGPT continue to make headways into hospitals and health systems, more organizations are realizing the importance of hiring chief AI officers and advisers, as they can help their enterprises align on AI strategy, governance, adoption and road maps, Dennis Chornenky, first chief artificial intelligence adviser of UC Davis Health, based in Sacramento, Calif., told Becker's.
Mr. Chornenky, who was appointed to the role of chief artificial intelligence adviser of the health system on June 21, previously held roles at the White House and UnitedHealth Group's Optum Health.
At Optum, he developed an AI governance framework and managed AI applications for clinical and business environments, and at the White House he served as senior adviser and presidential innovation fellow, where he focused on national AI policy.
Becker's spoke to Mr. Chornenky about his new role, the opportunity for AI advisers in healthcare and where he sees generative AI advancing in the industry.
Question: What are your top three priorities as chief AI adviser for UC Davis Health?
Dennis Chornenky: Currently, the first priority is AI governance. So I'm focused on building the guardrails to help ensure safety and regulatory compliance is top of mind for the applications that we're pursuing, and making sure that they are feasible and and are improving care delivery and patient-provider experiences.
The second is the AI adoption roadmap. I'm working closely with Ashish Atreja, MD, who's our CIO and chief digital health officer, to do a deep dive and assessment of the current state of our capabilities, technological maturity, and what the desired future state looks like to build an AI adoption road map.
Third, is education and training. Democratizing AI to help us discover new applications and training professionals to become comfortable with AI.
Q: What are some of the opportunities and challenges as being the first to be appointed to this role?
DC: I think more organizations are realizing the importance of a role like this, as it can help align an enterprise on strategy, governance, adoption road maps, these kinds of things. So I think that's the opportunity.
The challenge is focusing on accelerating things and moving them toward a higher state of maturity as well as bringing on more advanced technologies at a rapid pace. Generally, AI in healthcare is still a relatively new proposition, especially large language models and generative AI and their potential impact on patient-provider interactions, so there's a lot of need for mutual learning.
Q: ChatGPT and generative AI has been making quite a buzz in the healthcare community. Where do you see some of the biggest opportunities for this in healthcare?
DC: I think there is transformative potential. ChatGPT and generative AI are definitely powerful tools that we can use in healthcare, especially with patient communication and engagement.
We're getting to a point now where these technologies can actually work extremely well and they can really substantively improve interactions with patients by answering their queries, providing health advice and explaining information about their health records as well as generally helping them manage their conditions. These are all things that I think on the patient side are becoming more real and more possible than ever before.
Q: You are tasked with helping UC Davis Health develop a safe and ethical approach to AI. How will you approach this?
DC: At UC Davis Health, there is a health data oversight committee and some related subcommittees that have set up pretty good processes for reviewing these kinds of things and ensuring that the data and algorithms in use are being reviewed and deployed appropriately.
Now we're looking at AI governance. One is an emerging industry framework, referred to as responsible AI, which essentially helps to ensure regulatory and ethical standards, and it may include things like an AI review board or some kind of AI or analytics oversight committee.
Another aspect of governance is called model ops. Model ops helps you to monitor and manage the technical performance of these advanced technologies. Because we want to ensure that we know where the models are deployed, that they're behaving as expected, and if there's any kind of drift or unusual behaviors, we want those flagged and adequately addressed.
Q: How will AI shape the future of healthcare?
DC: It's gonna have a major impact on healthcare, and I'm already seeing that every day. Everybody's scrambling and a lot of boards are getting pressured and a lot of executives are wondering how they're going to adopt and scale this.
They're seeing the tremendous possibilities and now trying to wrap their minds around the best way to get these technologies up and running.
This technology has the ability to improve patient and provider experiences, it can improve system operations and efficiency, and I think it will play a large part in how health systems operate.