I have one word for you about artificial intelligence:

Relax.

The tsunami of artificial intelligence (AI) hype of late has me scratching my head. In 2022, ChatGPT rolled out with such fanfare that you’d think sentient, fast-footed robots would be due out next year.

Not that there isn’t exciting technology in development that will make us better at what we do. That’s certainly happening, and more about that later. But the reality is that we’re far from anything approaching true artificial intelligence.

Here’s the problem with “AI.”

The phrase “artificial intelligence” is, itself, misleading, if not completely off the mark. “Artificial intelligence” suggests brain power akin to that of humans. The title implies that AI can do what humans do. But it can’t and won’t until it can apply critical thinking.

Critical thinking is what makes our species unique and, indeed, supreme. With critical thinking, we apply an intellectually disciplined process of actively and skillfully conceptualizing, applying, analyzing, synthesizing, and evaluating information gathered or generated from observation, experience, reflection, reason, and communication. It’s the processes and information derived from critical thinking that formulate guideposts for our beliefs and actions.  

Frankly, when you start to think about AI against human intelligence and our capability for critical thinking, AI comes off kind of dumb.

Are machine learning systems (an AI subset), learning? Not in the human sense. Human learning requires self-regulation of multiple cognitive processes. Machine learning systems are simply accessing data and leveraging an increasingly deeper well of it to output a product based on preset rules (algorithms). Fancy technology? No doubt. Learning? Nope. There is no self-regulation of multiple cognitive processes in play, so maybe we need air quotes to describe machine “learning.”

Fortified with AI!

The concept of AI isn’t new, of course. Automated assembly line machinery, online search engines, smart phones, and many other everyday doodads utilize components commonly referred to as AI and have for some years. (By the way, the term “artificial intelligence” dates to 1956 and was coined by one of the founding fathers of AI, John McCarthy, an American computer scientist.)

But AI of late has become buzzy in the marketing landscape and increasingly murky in understanding. Nevertheless, many brands feel compelled to use it (the term) for fear of appearing technologically antiquated in the minds of their customers. AI as a marketing hook works because most consumers have little understanding of what it is, let alone how it works in their doodad.

Forces of resistance.

Here are two more reasons you need not worry about sentient, fast-footed robots at your door anytime soon.

1. Unwillingness for change. Historically, the mindset in the healthcare technology sector has not been one to demonstrate speed when it comes to adapting to change. Far from it. And the issue isn’t as much about incorporating AI (or any technology upgrade) as it is about flipping a mindset to one that welcomes change. This is how we think about technology at MediSprout. We are incorporating AI and AI-like technology where it makes sense now—automating routine and nonclinical functions, for example. More importantly, the foundation of our platform is built for growth that comes through change; change driven by new technologies and the expanding needs of patients and growing practices. 
2. Too few partnerships. Historically, technological change on a revolutionary level happens when two or more ideas evolve, often independently, and then merge. There is likely no better example of technological convergence or collaborative networking than the smart phone sitting on your desk. Change happens with partnerships. Industries rife with siloed organizations (I’m looking at you, healthcare) will be slow to innovate. Open and collaborative businesses will be all over it.

“What’s our AI strategy?”

There’s something else emerging from all the hype around AI lately: companies are feeling compelled to formulate an AI strategy. Personally, I think “What’s our AI strategy?” is the wrong question to ask. It’s too broad, unwieldy, and assumes understanding of AI in the first place.

A better question: “How does AI fit into our existing strategy?” This question prompts greater study of AI technology and pushes you to more specific considerations about how best to use it for your organization’s unique needs.  

There are many other questions to ask about artificial intelligence. It’s a big subject with big issues and even bigger consequences if we don’t ask the right questions. To wit:

• Do we need greater regulatory oversight?
• If we ask AI machinery to carry out specialist tasks, should it be licensed the way humans are sometimes required to be when carrying out specialized tasks?
• What of privacy, patient rights, and other ethical implications?
• How do we ensure that the human component is maintained, that is, retaining a workforce and not replacing it?
• Do we want machines making healthcare decisions?
• What other aspects of our current infrastructure need to be upgraded to support AI?
• Who is going to cover the investment needed?

Only the start.

We’ll have more to say about these and other issues of AI in the days ahead. Make no mistake, there’s plenty to be excited about. One thing AI can do better than humans is process enormous amounts of data, expeditiously. For healthcare, that’s life-savingly huge. AI’s capability to manage and accelerate massive flows of data through clinical trials is already helping drug companies deliver new treatments to the market faster. Generative AI is driving advances in medical imaging and speeding up processes that lead to faster diagnosis.

No one would disagree with the potential that AI holds to create better healthcare experiences for everybody. That said, let’s move ahead with incremental steps, better understanding, and less freneticism.

Ben Putland is the chief operating officer at MediSprout.