NYU Langone Health’s groundbreaking Generative AI Prompt-A-Thon brought together clinicians, educators, and researchers to harness AI’s potential in healthcare. The event tackled real-world healthcare challenges using large language models, emphasizing the need for human oversight. Dr. Jonathan Austrian initiated this event to bridge the gap for frontline healthcare professionals, fostering interdisciplinary collaboration. The results promise ongoing AI innovation, curriculum development, and technological insights. Austrian recommends similar events in healthcare, emphasizing mentorship and robust infrastructure, underscoring NYU Langone Health’s leadership in AI-driven healthcare transformation.
NYU Langone Health recently spearheaded an innovative initiative called the Generative AI Prompt-A-Thon in Health Care, which showcased its commitment to advancing the field of artificial intelligence in healthcare. In this transformative event, diverse teams comprising clinicians, educators, and researchers joined forces to harness the power of artificial intelligence, using real-world, anonymized patient data to tackle pressing healthcare challenges.
Hosted by NYU Langone Health’s MCIT Department of Health Informatics, Institute for Innovation in Medical Education, and Institute for Excellence in Health Equity, this inaugural Prompt-A-Thon brought together professionals from various disciplines. Their mission? To explore AI-driven solutions that could revolutionize healthcare, all while utilizing extensive patient data.
The event focused on large language models (LLMs), which are adept at predicting the next word in a sentence, paragraph, or essay based on vast contextual data collected from the internet. These generative AI models, sometimes referred to as LLM systems, employ a blend of likely next words to introduce variety and creativity into their output. One remarkable feature of these models is their proficiency in summarizing extensive texts, extracting vital information from databases, and engaging in human-like conversations as chatbots.
However, it’s essential to acknowledge that these AI programs do not possess genuine reasoning abilities and may generate conclusions or references that lack real-world validity. Hence, careful human oversight remains crucial, particularly in healthcare settings, where AI has the potential to enhance safety and elevate the quality of care.
Identifying the Challenge
Dr. Jonathan Austrian, Associate Chief Medical Information Officer for Inpatient Informatics at NYU Langone Health, explained the motivation behind this innovative Prompt-A-Thon. The challenge was to engage members of their workforce who possessed transformative ideas for generative AI but lacked the technical expertise to participate in other capacity-building initiatives, such as exploratory access or mentored projects.
While these existing programs successfully catered to highly motivated colleagues who could safely experiment with real clinical data or proprietary research concepts, they did not address the needs of frontline clinicians, researchers, educators, and operational leaders. These professionals comprehended the healthcare system’s challenges but required focused time and in-person support to bridge the gap between generative AI and their specific needs.
The Solution: Generative AI Prompt-A-Thon
The Generative AI Prompt-A-Thon in Health Care emerged as a dynamic solution to rapidly engage a broader segment of NYU Langone Health’s workforce in generative AI. Notably, it aimed to lower barriers to entry by offering an in-person event complete with at-elbow mentorship from generative AI experts, all without requiring prior experience with generative AI.
Dr. Austrian likened the event to a forum for interdisciplinary collaboration, a place where professionals from various specialties, roles, and backgrounds could come together to foster innovative ideas and relationships. This community of learning, he believed, held the potential to catalyze genuine innovation in healthcare.
Positive Outcomes and Future Directions
The results of the Generative AI Prompt-A-Thon promise to have a lasting impact on NYU Langone Health. Firstly, the participants, both in-person and remote attendees, have now become a part of a growing community of learners with ongoing access to GPT technology, updates on emerging AI solutions, and additional avenues for leveraging generative AI.
Secondly, the event generated a plethora of innovative ideas that will likely be refined by the community and eventually transformed into practical applications within NYU Langone Health and shared with the broader healthcare community.
Thirdly, insights gained from mentors and survey data will inform the continued development of NYU Langone Health’s internal capacity to harness generative AI.
Moreover, NYU Langone Health is expanding its collaboration with health science library staff to formalize a curriculum in generative AI, building upon the success of this event. They plan to take Prompt-A-Thons on the road, engaging smaller groups of researchers, educators, clinicians, and corporate services members.
Lastly, the event provided valuable lessons on the technological infrastructure needed to support scalable and intensive generative AI usage. Partnerships with organizations like Microsoft were essential to ensure a seamless technological experience and pave the way for future expansions.
Recommendations
Dr. Austrian recommends that other healthcare organizations consider organizing events similar to the Generative AI Prompt-A-Thon in Health Care. He advises dedicating more time to explaining the fundamentals of prompt engineering, ensuring a balanced mentor-to-team ratio, and establishing a robust technological infrastructure to support such events effectively.
Overall, NYU Langone Health’s Generative AI Prompt-A-Thon in Health Care exemplifies a forward-thinking approach to fostering innovation and collaboration in healthcare through the power of generative AI. It not only engaged a diverse workforce but also positioned NYU Langone Health at the forefront of AI-driven healthcare transformation.
