Anyone tuned into social media during the recent global pandemic may remember clicking on links, pictures, and videos from talking heads discussing wild accusations and ideas about all kinds of information.
From the origin of the virus to the many home remedies sure to cure even the slightest sniffle, we were overrun with a fantasy-footballish level of misinformation that wasn’t going to do anyone any good.
The Joint Research Center (JRC) from the EU recently published a fantastic report about disinformation and misinformation that emerged during the pandemic.
In most instances, this revolved around reshaping content around political and societal topics, allowing people to share information on a global scale that supported misleading or false claims targeting policy actions.
Now we operate in a new era between political campaigns when the age of artificial intelligence (AI) stands on the precipice of transforming our current information standard. That is especially relevant in the world of healthcare.
The Good Side of AI-Supported Healthcare
Let’s get the good news out of the way first. Yes, AI can make dramatic improvements to our U.S. healthcare system.
We have a current model that relies on private funding and growing price tags, and any system that can streamline operations, improve patient care, and reduce worker stress is going to be beneficial.
We are likely to see AI implemented in healthcare through:
Telemedicine: The pandemic taught us that virtual healthcare is not only a possibility, but a much-desired reality. Johns Hopkins Medical Center put it best.
With AI-backed telemedicine, we will see improvements in palliative and patient care, greater control of infectious illness, better assessment of needs, and support great family/provider connections.
Productivity & Comprehension Enhancements: We will see a revolution in healthcare efficiency. Over 83% of executives managing hospitals, care centers, and group homes agree AI capabilities will address health-related challenges worldwide. This could be anything from immersive MRI scans to preventative health screenings.
Inclusive Diagnosing & Reporting: AI doesn’t need to sleep. Whenever patient information is entered into a digital system, it can be scanned for keywords, phrases, medications, and treatment plans, then pull out insights that our brains are not capable of seeing. This will reduce potential caregiver/provider errors and lead to better patient outcomes.
A Re-Balancing of Worker Time: Around 70% of healthcare worker tasks will be reinvented by technology augmentation like AI-backed automation.
This will transform the ability of caregivers from writing background notes and spending a lot of time researching to focusing on patient needs and getting to know clients at a more personal level.
This could uncover insights and trends that would never have been possible before when we have such a turnkey operation.
AI Healthcare Concerns
Now for the other side of the coin. While we certainly can see many benefits from introducing AI tech into healthcare systems, we do need to proceed with caution. AI is not infallible. It can make mistakes because it is supported by human users.
These systems learn to make decisions based on training data and input information. That info is subject to human choices that reflect historical or social inequalities.
To put it simply, if a doctor using an AI tool or a programmer creating the tech already has racial leanings, gender-based bias, or some other form of social injustice ingrained into their minds, they are likely to perceive and enforce those same ideals with AI.
How did Uncle Ben from Spiderman put it best? “With great power comes great responsibility.”
Yes, there are click-bait articles about how 42% of CEOs say AI could destroy humanity in the next 5 or 10 years. You should take any kind of claim like that with a grain of salt. What is better to pay attention to is where the money is going.
The White House has called on silicon valley and the bigwigs of AI to gather together and help form a new regulatory roadmap to ensure we can leverage this technology for the betterment of society and not for outlandish use.
That may help in the long term, but there is a greater issue related to healthcare and AI – misinformation.
The Misinformation of Healthcare AI
For years we have seen warnings from top engineers, software developers, and tech angel investors that AI poses a real threat to the value of information. We are leaving the golden age of information and entering the age of verification.
Everything from synthetic images of political ad campaigns to how eating a raw lemon will cause you to experience rapid weight loss is only going to get louder and more convincing.
Tools like ChatGPT, Midjourney, and more are able to recreate information in such captivating ways that you would never know their authenticity.
The risks associated with using AI in medicine and general healthcare can cause patient harm due to breaches of data privacy and security, as well as social inequalities.
For example, using an AI-driven pulse oximeter that overestimates blood oxygen levels in patients with darker skin can result in the undertreatment of hypoxia.
But the risks don’t stop in the ER. There is also public discourse on and off social media. AI-generated content risks intensifying issues we already have and making it far more challenging for users to best interpret the world around them.
Remember, this is a “feed-based system.” It relies on what works and what we give it for base data.
So, if an internet troll in someone’s basement devises a strategy for shooting up the TikTok ranks by spreading unverifiable “facts” about how to treat disease, stop the spread of viruses, improve sexual health, or even what foods to feed our babies, what is to stop them?
AI makes content so convincing that even the most well-intentioned mother with a nursing degree is likely to make a mistake due to the inundation of force-fed content.
A Brighter AI Future
Don’t worry. AI isn’t all doom and gloom. Like a teenager attending the first party with parents stolen alcohol, we need to go into AI with eyes wide open.
Now is the time for healthcare workers, medical institutions, tech gurus, and political leaders to gather and start hammering out some solid guideposts.
Whether that is regulation, auditing systems, or some form of verifiable warning system, we need to ensure our data is protected and more emphasis is placed on scientific method-backed results.
Together, we can still leverage the power of artificial intelligence to improve healthcare norms. It is going to take some trial and error and getting past the fear of change so we can welcome a new era of health, enlightenment, and, God forbid, human decency in the online sphere.
Written by: Emmanuel J. Osemota
