This was a study by Dr. Pearl Subramanian and colleagues, which really asked the question of what information are patients getting online when they ask about AI and cancer care? So not about using AI to give them information about how to treat their cancer, but if they want to know what can AI do in cancer care, what information are they getting? And they looked at both Google and YouTube, so Google web pages, YouTube videos...
This was a study by Dr. Pearl Subramanian and colleagues, which really asked the question of what information are patients getting online when they ask about AI and cancer care? So not about using AI to give them information about how to treat their cancer, but if they want to know what can AI do in cancer care, what information are they getting? And they looked at both Google and YouTube, so Google web pages, YouTube videos. This was not chatbot-generated content, but really indexed web content. And they found a really small number of resources, both on Google and YouTube, that address this. And most of the resources were at about a 12th grade reading level, which is higher than the average reading level in America. And most of them did not have good quality information. And notably, less than 20% address the safety risks of AI with hallucination and misinformation. And that’s really significant because we need our patients to understand what are the limitations of AI. And this really speaks to the need for broader patient guidelines and content about what AI can do, but also what AI can’t do and the potential dangers of AI.
This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.
