What is AI and why is it hot right now?
Since its release to the public at the end of 2022, everyone’s been talking about ChatGPT, an AI chatbot based on the Generative Pre-Trained Transformer (GPT) language model. In short, ChatGPT is a convenient conversational interface that allows users to ask questions and obtain AI-generated answers in natural language. Our recent research shows how AI can impact several societal megatrends; my previous blog addresses one of those megatrends: digitalization.
Although ChatGPT is sparking huge media coverage, it’s just a single use case for AI, and a minor one at that. In other words, it’s just the tip of the AI iceberg.
For several years, technologists have been thinking about ways to embed AI in products and services to improve the customer experience. We’re just beginning to understand how we can exploit its full potential in a wide variety of industries and societal fields and how AI can reshape the way we see the world.
AI and healthcare
Healthcare includes both preventive and curative care, but since its beginnings, the medical profession has focused more on curative care—making people healthy again. In the last decade, also several AI applications have begun to focus on curative care, especially detecting disease patterns in images (e.g., specific cancers on radiology images) and providing decision support systems.
A more concerted effort to focus on preventive care, namely in Public Health, can have a huge impact: Every dollar spent on preventive care will pay us back by reducing the need for curative care and overall healthcare costs. However, the public health sector has a long way to go in leveraging AI’s huge potential to improve global health.
In our most recent peer-reviewed article, published in March 2023 in the respected MDPI publication “International Journal of Environmental Research and Public Health,” my co-author Daniela Haluza and I investigated AI’s potential for contributing to the public health sector. The article is based on a study in which we used OpenAI's recent GPT model, a more robust version of ChatGPT, to analyze AI’s potential for contributing to and re-shaping societal megatrends, including the challenge of improving global health. Our study concludes that AI can significantly improve the medical profession’s understanding of several solution areas and lists concrete use cases for new AI-based solutions.
Health care personalization
The study shows that generative AIs such as GPT can summarize huge amounts of public health data and evidence-based literature to create on-demand literature reviews that are generated in real time. The automation of public health data summaries can give us a better understanding of the current state of public health and inform decision-making and policies that will improve health outcomes.
In addition, these summaries can easily be written in the form of news stories disseminated to the public. AI chatbot natural language processing (NLP) algorithms can process and analyze large amounts of patient demographic information, lifestyle habits and other contextual factors to identify individual health risks, generate personalized health messages and recommend interventions tailored to the individual’s health needs and preferences.
Predictive future outcome modeling
AI simulations and forecasts allow governments and the medical profession to explore the impact of public health policies on a variety of demographics and compare the effectiveness of various strategies and interventions. This can help them make more informed decisions on how to manage and improve public health.
AI technology can develop predictive models of public health outcomes by collecting and analyzing huge amounts of data from various sources, such as patient records (including age, sex and lifestyle) and social media (including dietary habits and physical activity). For example, AI chatbot systems could predict the specific risk of Spanish-speaking individuals developing type 2 diabetes with high accuracy.
By analyzing this data and determinants of health, AI systems can identify trends and patterns in health-related outcomes and provide personalized suggestions and automated decision support for clinicians that allows them to make more informed decisions with greater accuracy and speed, improve resource allocation and enhance treatment protocols.
Detecting and preventing public health fraud and abuse
Currently, detecting and preventing public health fraud and abuse is a challenging exercise. AI systems can help by scanning large amounts of data in financial and medical records and detecting suspicious anomalies and irregular activity such as:
- An unusually high number of hospital visits in a short period of time
- An increase in the length of hospital stays in times of lower demand for hospital beds
- Over-prescription and/or larger doses of medications
- Fraudulent business generation funnels
The other side of the coin
Every technology comes with risk, and AI probably carries an exponentially higher risk than earlier technologies. For instance, only 54% of AI projects make it from pilot to production. Other risks include AI’s potential to perpetuate or even increase inequalities among demographics, widen income gaps, improve access to healthcare for the wealthy and reduce access to people in need.
To address global health challenges, governments and the medical profession need to use AI wisely. The immediate need for discourse about AI’s contributions to scientific research is clear. The discourse should update and standardize policies, establish systems, and build guardrails that advance society’s best interests and prevent malicious actors from abusing AI for their own gain at the expense of others.
AI is here to stay
There’s no question that AI can revolutionize our approach to global health by predicting the likelihood of disease in certain demographics, identifying and monitoring emerging public health threats, preventing disease via dissemination of relevant population-specific information, supporting research, facilitating data-driven diagnosis and decision-making, driving cost savings and preventing abuse and fraud.
Anyone interested in learning more about AI and its potential influence on the other megatrends is invited to read our first study here . Stay tuned for more blog posts addressing AI’s contributions to additional areas of our lives.