DailyPost 3059
HEALTHCARE – AGENTIC AI FOR PATIENTS
There are many IT dreams that crack up completely or partially when the rubber meets the road. Not that they are misconceived or the capacity to initiate confidently is not there, what is generally missing is a level of superhuman persistence and the capability to prove its immense utility to the world as a proof concept. One of the very recent dreams which has certainly failed currently has been the idea of Metaverse. But Artificial Intelligence is no such dream, it has been the field of intense research since 1956 and started yielding clear cut results since the days of DeepBlue.
DeepBlue in 1997 defeating a reigning world chess champion to the advent of ChatGPT towards the end of 2022, taking the world by storm in no time, is the momentous history of AI in the modern age. AI graduated in a very short time of its public advent to become synonymous with IT. AI is turning out to be the biggest game changer after the Internet. Generative AI became the order of the day. Generative AI to Agentic AI has been the next great leap. What is Agentic AI? “These are autonomous digital agents capable of understanding tasks, planning, and executing multistep processes, effectively acting as digital workers.”
In his keynote at CES 2025, Jensen Huang, NVIDIA CEO, predicted that 2025 will be pivotal for the deployment of AI agents. So the generic LLM or even SLM can be brought further down to a use case of that enterprise, task, workflow, customer base, type or in case of healthcare patients. Healthcare can have any number of use cases given the multifarious functions, interfaces and the deliverables. From radiology to medication can be the gamut of complex tasks of a variety of AI agents. What we are discussing today is how we can create an AI agent which can act as a one stop shop for patients.
How much of our medical records are available with us? That too in a usable form. Whatever little we have, we keep struggling with it? Can we make head or tail out of an MRI scan or ultrasound films? Can we connect the MRI report to the prescription? Do we have any idea of how insurance claims have been settled? Why has something not been approved? Do we even remember data, time, medical interventions during the whole duration of one treatment? I am talking of one treatment span, think of all the treatments in a decade or two or more. My facts and data are unknown to me, even if we carry some with us, including the ones on a digital medium. We are kept bereft of medical intelligence about us, if it were only because of lack of technology, technology is right here. If it involves other issues as well, it needs to be ironed out sooner than later, the time has come. We need to be active participants in our health journey.
AGENT AI FOR PATIENTS CAN REVOLUTIONISE OUR UNDERSTANDING OF THE HEALTHCARE BEING PROVIDED TO US.
Sanjay Sahay