Ensuring AI is developed without bias in healthcare
As part of our ongoing discussion series, Under the Same Sky, Abner Mason, founder and CEO of SameSky Health, met with met with Kevin Dedner, MPH, author, entrepreneur, and thought leader. Kevin described his experience and lessons learned in solving complex public health problems over the past 25 years as a public health executive. He also discussed his current project, where he is challenging the bias in healthcare as the use of artificial intelligence (AI) continues to rise. Kevin is working to launch an AI-powered platform to provide trustworthy health information, referrals to care, and care support to traditionally underserved communities and priority populations to decrease health disparities and advance health equity. View the recording.
The emerging conversation around AI has taken over the healthcare industry, but what’s missing from the conversation is the importance of health equity and improving health outcomes for underserved populations. Kevin believes we need to push back on the narrative that AI will be bad for people of color and underserved populations. AI is all about the inputs, so if we are thoughtful and careful about the information we are feeding in, and are vigilant to not include bias, we could use AI to improve health outcomes.
“Don’t run away from the potential of AI, but instead embrace it!””
We often see that people of color are misdiagnosed or diagnosed very late in the stage of whatever condition they might have. Kevin’s vision is that if we build a trusted platform where people can get good information instead of misinformation, it will speed up their time to diagnosis, which would also be a great outcome for payers. AI could give people articles, video content, or even conversation guides to help them go back to their physician and better explain their situation. Once a proper diagnosis is given, this platform would be able to support them through their care journey and help them continue on a path with correct information and resources. Kevin thinks AI shouldn’t necessarily be in the business of diagnosing people, but instead give people the tools that they need to get diagnosed much sooner.
Our conversation with Kevin raised a number of important issues that deserve consideration. Abner and Kevin agreed that there are always problems to solve when trying to make a difference, but the number one focus when building new technology needs to be recruiting the right people. Will people be reflective? Will they challenge their own assumptions? Ask these questions when building a team because this is a big part in how AI works and challenges the assumptions in the data. People will be your biggest and best resource when building AI technology.
Overall, we want to reduce health disparities and achieve health equity for underserved populations. Technology and data sources need to have these populations in mind. Kevin suggested that most of today’s AI data is based on the majority populations, so we need to over-sample the underserved populations to make sure we’re getting a good picture of what health conditions could look like for all the people we intend to serve.
If we don’t put together teams that are challenging themselves, we will see the bias that has persisted in the healthcare industry continue. Kevin mentioned that there are people running from AI technology because they are worried it will widen the gap of health disparities and be bad for people of color. The theory is that AI will only perpetuate these current biases; but no, there is an opportunity to change that. If we build a thoughtful, reflective, and diverse team that is very deliberate about the data we are using to drive algorithms and decisions, we can actually close the gap in health disparities.
“This is an opportunity to embrace the technology and do something incredibly different. It’s important for entrepreneurs to see AI as an opportunity instead of something we should be afraid of. Move forward optimistically but cautiously.””
Related
