ALREADY? WE JUST DISCUSSED ARTIFICIAL INTELLIGENCE A FEW MONTHS AGO
Yes, we did, and yes, that previous discussion is already incorrect on the timeline. Artificial intelligence (AI) was thought to be upon us in everyday life on a 3- to 10-year horizon. Well, as it turns out, 3 to 10 months is more like it. Microsoft just released a version of AI enablement called Copilot, a package of AI services likely to infiltrate a plurality of corporate offices and small businesses alike by the end of 2025. Currently, it includes all sorts of ways to increase efficiency in office work, such as creating a readout of an online meeting, searching for files more intelligently, and generating new content by instructing Copilot rather than authoring prose. We may still be a few years (decades?) away from a sentient computer that thinks for itself, but given the near lack of impediments to exponential growth of nonbiological neural networks that are so fast and complex that they need dedicated energy from a nuclear facility (you read that correctly), I believe that computer consciousness will exist in my lifetime.
AI BASICS THROUGH THE LENS OF PHARMACY PRACTICE
You might be surprised to learn that AI’s antecedents have existed for nearly three-quarters of a century. In 1956, Allen Newell, Herbert Simon, and JC Shaw developed Logic Theorist, the first artificially intelligent computer program.1 We’ve had evolutions of AI ever since in ways you may not recognize but that can help health care providers identify and solve problems more efficiently and effectively.
About the Author
Troy Trygstad, PharmD, PhD, MBA, is the executive director of CPESN USA, a clinically integrated network of more than 3500 participating pharmacies. He received his PharmD and MBA degrees from Drake University and a PhD in pharmaceutical outcomes and policy from the University of North Carolina. He has recently served on the board of directors for the Pharmacy Quality Alliance and the American Pharmacists Association Foundation. He also proudly practiced in community pharmacies across the state of North Carolina for 17 years.
For instance, algorithms are already used for drug utilization review, such as every time a pharmacist selects “refill too soon.” Patients are already using natural language processing to ask Siri or Alexa whether they can take a certain drug while pregnant or to inquire about the best drug for weight loss. Natural language processing is what allows us to use our human language to translate those words into a query or command, such as instructing a computer to water the lawn if it doesn’t rain later today.
Most inventory systems have some level of machine learning that adjusts orders based on historical results, new orders, and trends. You can imagine how valuable this will become for ordering medications if real-time inputs such as prescribing and pickup data, new manufacturer entrants, and incidence data about various conditions become part of the inputs and algorithms. There are already versions of AI being applied to pricing data on both the buy and sell sides in the pharmacy industry today.
Finally, generative AI is where things really get exciting, creepy, or scary, depending on your predilection for change and handing over decision-making and consciousness to nonhuman, nonbiological beings. Generative AI creates new content, not simply results from a complex set of algorithms. It won’t be long before pharmacists start asking a chatbot what the best approach is for a counseling session with the patient next in line. In fact, it’s likely already happening without the permission of corporate headquarters via smartphones. A personal pharmacist assistant is coming very soon. I believe Microsoft will be the company that pierces the veil mostly because nearly every business on the planet uses it and if Microsoft enables it, it becomes mainstream.
HOW MIGHT AI BE APPLIED TO HEALTH CARE AND PHARMACY IN THE NEAR FUTURE
Imagine a radiologist reviewing an image. Humans are OK at best, but a computer can analyze images with a level of spatial recognition that humans cannot touch. We now rely on this technology to open our phones or pass security at the airport, so what if 100 images could generate hundreds of probabilities with a problem condition list associated with each? The radiologist could focus their attention and judgment on certain diagnoses and images with higher probabilities. In this light, you can see how quickly this could be applied to pharmacy practice and discovering drug-drug interactions; optimal dosing; and individualized regimen building, counseling, and monitoring. The only limitations are data inputs, processing speed, and a human to assist until such time when a computer can exhibit judgment and fewer hallucinations (an AI term for obvious errors akin to your preschooler thinking the world is flat from all observable data).
WHY NOW?
In short, this acceleration is thanks to the explosion of data storage, internet data sharing, the contribution of data via the internet and human interactions such as social media and, vitally, processing chips. All are now scaled for the first time in ways 40-somethings like me would never have dreamed of when we were growing up with floppy disks. The ideas, methods, and implementations of AI have been around for decades, but things like Siri, Alexa, Facebook, Tesla (you’d be surprised by what data your vehicle collects), 1 GB bandwidth to hundreds of millions of homes, and 5G connectivity were the blood that needed to course through AI’s veins to grow into something we are only beginning to imagine.
WHAT YOU CAN DO NOW TO ADVANCE, NAVIGATE, OR EVEN JUST KEEP YOUR JOB
Search the internet for articles. Take a class at your local community college. Befriend an engineer. Start a club that reviews each member’s use of AI over the past week instead of a book. Have your kids teach you how to use ChatGPT or other chatbots. Press your associations to come up with continuing education or position papers on AI as the American Medical Association did earlier this year.2 Encourage schools of pharmacy to educate faculty, students, and residents about the emerging opportunities, risks, and necessary skills associated with AI. In short, don’t be passive: Get in the game or get left behind.
REFERENCES
1. Monahan J. Artificial intelligence, explained. Carnegie Mellon University Heinz College. July 2023. Accessed November 26, 2024. https://www.heinz.cmu.edu/media/2023/July/artificial-intelligence-explained
2. AMA future of health: the emerging landscape of augmented intelligence in health care. American Medical Association. February 26, 2024. Accessed November 26, 2024. https://www.ama-assn.org/practice-management/digital/ama-future-health-emerging-landscape-augmented-intelligence-health-care
