Tony Yi, chief technology innovation officer at MedVanta, a musculoskeletal management services organization, connected with Becker's to discuss how he anticipates ChatGPT and other AI programs will evolve in healthcare.
Note: This response has been lightly edited for length and clarity.
Tony Yi: Academically and professionally, I've studied and applied many hyped, disruptive and emerging technologies, but AI is in a class of its own. Its accessibility, reach and ability to make an impact in very basic and highly complex use cases are astounding.
To understand how the ChatGPTs and the Bards of the world will evolve, the focus should be more on how healthcare organizations will evolve around them. We also must make the distinction that AI is very broad, so most of my references will be toward chatbots utilizing generative and natural language processing technologies.
One of the first ways chat-based AI will evolve is in its ability to integrate with other solution stacks. This reminds me of the announcement made by Microsoft's Nuance Communications about their integration with OpenAI's GPT-4. Having the world's largest medical dictation data set coupled with ambient AI and now GPT-4 should be revolutionary. Another example of this is the integration with robotics process automation (RPA). It was announced a few weeks ago by Automation Anywhere that GPT/generative AI is now embedded into their Automation Success Platform. These examples show how the integration of different technologies makes them even more powerful. I see healthcare companies big and small racing to not just adopt, but to integrate with what's already in place.
Another way this evolves is how it will transform the delivery of care through its ability to go beyond traditional settings. Through digital engagement, accurate data and diagnoses, telemedicine, remote patient monitoring/remote therapeutic monitoring, Internet of Things, etc., advancements in AI can impact care more accurately. For population health and risk stratification, having a technology that can comprehend data, generate data, and proactively communicate more specific and targeted messages for patients with specific risk scores should improve the management and outcomes, at scale. A significant threat to us today is mental and behavioral health. For use cases, imagine an often-overlooked population in geriatrics. What COVID-19 taught us was that of all age groups, older adults benefitted the most from technology. Think Amazon's Alexa and other voice and remote technologies that facilitated and cultivated communication with loved ones during periods of lockdowns and isolation. The companionship, interaction, and the ability to connect was very powerful, not just with others, but for content, entertainment, meals, activity and lifestyles. I see AI more than just conversations with consumers; I see it as essential, and as a human-like modality.
Operationally, many use AI for blocking and tackling in a more transactional way, but as organizations become more understanding and trusting of its capabilities, its use cases will soar. From simple chatbot functions for queries, appointments, surveys, to full blown marketing campaigns and new content creation for legal, regulatory and human resources, AI will go from complimentary to a must-have internal capability. Having also managed large hospital systems and the IT operations required to manage them, I can also see how this evolves into the centerpiece for incident, problem management, service, risk and change management. Having longitudinal data with real time data should also empower IT leaders to reduce incidents and plan for potential disruptions and performance degradation. It should also be used for customer interactions due to its ability to analyze data and generatively respond, and its accuracy and capabilities will only get better. Potentially, you could even reduce and convert higher-level tier two incidents and problems to tier zero self-help. These scenarios highlight the ability for organizations to improve user experience and satisfaction scores on so many strategic and operational levels.
In the same ways that COVID-19 accelerated healthcare services and societal norms, AI-driven chatbots have done the same and will continue to directly impact healthcare services and delivery.