# Can generative artificial intelligence enhance evidence-based and personalized medicine?

## TL;DR

This paper explores whether generative AI should follow strict guidelines, offer personalized care, or find a balance in clinical decision-making.

## Contribution

The paper introduces a discussion on the optimal design of GAI for clinical applications, emphasizing context and personalization.

## Key findings

- A recent study found differences between GAI and physician treatment recommendations.
- Context and patient-specific factors are crucial in AI-driven clinical decisions.
- The paper advocates for a balanced approach in GAI design for better medical outcomes.

## Abstract

Should clinical applications of generative artificial intelligence (GAI) be designed to follow treatment guidelines rigidly, provide individualized recommendations, or be somewhere in-between? A recent study in PLOS Medicine comparing GAI versus physician treatment recommendations highlights the importance of context- and patient-specific circumstances in decision-making.

In this Perspective, Arun Thirunavukarasu discusses a recent PLOS Medicine study comparing generative artificial intelligence (GAI) treatment recommendations with physicians', raising the question of whether GAI should be designed to follow treatment guidelines rigidly, provide individualized recommendations, or be somewhere in between.