Agentic ai is a hot topic in numerous industries, including medical care. Everyone seems to be buzzing about AGI, imagining Jarvis de Iron Man, where AI systems are capable of human intelligence and decision -making capabilities.
The emotion is understandable: suddenly there is a real opportunity to build models and robots that imitate human tasks and that can interact and interact with humans. Companies are excited about this capacity, especially in industries with labor shortages to replace human workers with AI machines that can automatically tasks more efficiently, without errors or financial compensation.
But when it comes to medical care, I think we must take a minute to evaluate at a deeper level. In general, medical care is an industry highly regulated by a good reason and cannot assume some of the risks and failures inherent to AI that other industries can. We need to evaluate the AFFEE through the specific lens of the health industry and their workflows, evaluating the needs of suppliers, medical care organizations and patients, recognizing that medical care is not monolithic and will have an AI of workflow.
Understanding this requires being honest about the strengths and weaknesses of the Agent AI. Llms Excel In creative processes: for example, Chatgpt can instantly create a new song about Taylor Swift’s latest album in Dr. Seuss’s style with minimal indications. Comparatively, rules -based engines are good at structured output. Consider a notice for a driver without a car: “When a traffic light is red, then stop.” The fascinating thing about the Agent AI is that it will probably fall somewhere in the medium based on rules and partly creative, and navigate that average term and the appropriate risk and reward when the workflows of workflows of workflows of workflows would apply work flows Workflows Workflows Workflows Workflows.
The most useful heuristics that I have found in this issue is to analyze the concept of risk versus consequence. I define the risk as the percentage that something would fail and consequence as the result of that failure.
In scenarios in which the workflow risks are high, he does not want the agent process to possess it, the reality is that each model of AI will fail at some point and the cost of that fault could be high when it comes to results of medical care.
Here are two examples in which agent workflows would not work:
- Authorning or defining advanced directives (planning at the end of life). Clearly, there is a creative and interpretive workflow here requesting empathy, human experience and judging when to guide when listening; Because the source of information (people and their situations) are not all the same. It is also a situation with high consequences that does not want to make mistakes in any way.
- Triaje management in a chaotic and fast environment. People are the most suitable for making quick decisions: there is no time to enter data into an agent.
However, here there are two examples or where the agent would work in medical attention:
- EHR data unlock using an agent for an automatic task sequence that requires sailing a user interface. The business level software has done it before. It used to be called RPA or automation of robotic processes, but now the processes with agents can do this with much more resistance.
- Review patient graphics to ensure that emerging chronic conditions were not surprised by doctors.
At this time, I believe that Agentic AI is at a stage in which he will fail if he begins to tell doctors what to do and take on decision making, and what is worse, harms everyone’s confidence (for his association) with artificial intelligence in medicine. When patient safety, empathy and human judgment have priority over cost savings and possible efficiency improvements, a human must be aware. But automatize tasks, take notes, extract data, the tedious manual processes in medical care that do not require human intervention in each step, is a great place to start implementing AI. The Healthcare Industry Should Be Looking at How Ai Chan Find and present Care Care Care Care Care Care Care Care Care Care Care Care to Care To Outtcomes Care.
Photo: Filo, Getty Images
Isaac Park spent his youth playing with technology and in high school, his formal education began in software development. After moving to Durham, North Carolina, graduated from the University of Duke with a degree in computer science. Isaac is supplicating his technological career immediately, working as a software developer, building front frames. Subsequently, he moved to a product management role, guiding interested parties and technical teams through a wide variety of projects, from the beginning to the final launch.
In 2009, I have co-facounded an innovation and product studio, pathos ethos, and guided startups and corporate innovation teams through business-changing digital products in the Healthcare and Defense verticals: All The Way From Releaseing A Multi-Million Delivery, To Bilding Ysed Use, To Buges Dollar Range Dollar Range, To Buges Dollar Range, To Bilding Ysed Use, To Bilding Dollar Dollar Dollar, Competitive. At the end of 2022, the ethos of successful pathos and joined the Duke Pratt School of Engineering, serving as a faculty for the center of the Christensen family for innovation in product management and innovation. In 2023, he co -founded a Native Health Technology Company, Keebler Health, and currently works as CEO.
This publication appears through Medical influencers program. Anyone can publish their perspective on business and innovation in medical care in Medcity News through influential people of Medcy. Click here to find out how.
About The Author
