We have access to more health information than ever before - so why can it be so difficult to make decisions about our health, such as which medication to take or which procedure to undergo? Information overload aside, our biases and habits can influence our decisions in ways we don’t realize.
Dr. Trent Cash, a Lupina Foundation Postdoctoral Research Scholar based in the School of Psychology, is investigating how self-awareness of our decision-making processes (known as metacognitive knowledge) can lead to better decisions about our health and beyond. And yes, that can include using AI chatbots. Dr. Cash explains:
Why is it important to understand our decision-making processes better when it comes to our health?
Medical decision-making is hard. When we make these decisions, we have lots of information and options to consider. To make decisions that align with our priorities, we need to have a good understanding of who we are, what we value, and – sometimes – what we’re willing to tolerate. For example, imagine you’re picking between two procedures. One is painful and has a short recovery time, while the other is less painful, but has a longer recovery time. To pick the option that will make you happiest, you need to know whether the recovery time or the degree of pain is more important to you.
This kind of self-insight is also important because it allows you to communicate with healthcare professionals. If you can’t tell the doctor what is important to you, then it will make it difficult for them to guide you in the right direction.
How can AI help us better understand how we’re making decisions?
AI chatbots – such as ChatGPT and Claude – are incredibly powerful tools, and they’re getting smarter every day. I believe that there are two essential ways that AI chatbots can help us make decisions. First, they can fill in gaps in our factual knowledge. For example, imagine that you are deciding which medication to take for a newly diagnosed condition. The chatbot can easily identify information about each medication, such as what drug class it comes from, what other medications interact with it, and how much it costs. In this sense, AI chatbots can function like a more-powerful version of Google search.
Second, AI chatbots can be highly effective as sounding boards. You can tell the chatbot about your feelings, ask it to challenge your beliefs with counterfactuals, or simply tell it what your priorities are and ask it what it would do. While I would never encourage someone to take the chatbot’s opinions as definitive truth – they can definitely be wrong – they are often quite skilled at summarizing information and engaging in reflective conversation. In this sense, I like to think of the chatbots as journals that can respond to you and help you work through your thoughts.
What should we watch out for if using an AI chatbot to help make health decisions?
Despite their many strengths, AI chatbots are still just tools with a very friendly and confident voice. They are not experts, and they do not have human-like experiences. While they can help you think through a decision, they should not be considered a replacement for consulting a medical professional – and they definitely should not be allowed to make a decision on your behalf.
Think of an AI chatbot as a very smart stranger that you met on the street who knows a lot of facts. Would you let that stranger choose what medication you took? If not, then I wouldn’t let a chatbot make that decision either.