*A brief, non-technical explanation. (Like one you might give a curious friend at a dinner party.)
Clarke’s Third Law: Any sufficiently advanced technology is indistinguishable from magic.
Arthur C. Clarke, Profiles of the Future (revised edition, 1973)
Artificial Intelligence (AI) is the next big revolution in computing. You interact with AI already when you use Google maps, ask Alexa a question, click a recommendation on Netflix, or book an Uber. Like previous revolutions (e.g., PCs, internet, mobile technology) it will affect every aspect of business and life.
What is AI and why now?
People don’t agree on what AI is. Computer scientists in the 1950s first set out to create a human-like intelligence. This inspired visions of robots taking over the world: as in 2001: A Space Odyssey, the Matrix, and the Terminator movies. Right now, it is more useful to think about AI as a sudden leap in computers’ ability to solve specific problems. Two developments have driven this leap: a rapid increase in cheap computing power and an explosion of digitized data. The power comes from new computer processors, (originally designed for video gaming), and the rise of inexpensive cloud computing. The data comes from the trillions of machine interactions we have every minute. AI involves computers crunching enormous amounts of data very fast, over and over to program themselves. The theory has been around for more than half a century, the ability to do it is new and moving at an incredible pace.
How should healthcare leaders think about AI?
There is a lot of confusing AI terminology and many ways to categorize its capabilities. It’s not critical to know all the techniques, but it is important to know where to use it. Healthcare leaders should think about AI in two ways: automation, and insight generation.
Automation is the ability of computers to take over certain tasks. This can be either at the same level as a human, or in superhuman ways. Like other technologies today, AI automation will either support humans or replace them. For example, an AI tool could be “trained” to process CT scans and flag anomalies for a radiologist to review. AI tools could protect systems from cyber-breaches, pull up ED patient records on arrival, or translate for a patient’s family. Doing things humans could do, often faster, more reliably and without getting tired.
Insight generation is the ability to analyze data and predict the future. Computers will be able to spot patterns in large data sets and create new insights that humans cannot. In medicine that means discovering new disease-markers, better care models, and new treatments. In health care management it means better strategic and operational decisions.
What will AI mean for healthcare?
Right now, very little AI is used in clinical practice, but that will change rapidly. Specialties already digitized, like radiology, pathology and telemedicine will use AI first. Health care operations will also benefit given large amounts of data coming from new IT systems. Where we currently collect enough data in a machine readable format – AI will be able to use it.
For patients, AI will mean better treatments and more standardized outcomes. In medicine, this will lead to more rapid development of new drugs, techniques and cures. In healthcare, AI will improve the chances of getting the right treatment at the right time. Good outcomes often depend on luck. Are you in a rural hospital with a neurologist to identify stroke? Did your specialist randomly read an article yesterday that helped her diagnose a rare disease variant? Did your care team use the most agreed treatment? an optimal dose of the medication? Is the nurse having a bad day? AI promises to raise the knowledge and speed of every healthcare provider to the highest level.
What are some of the challenges of AI?
AI comes with many technical, cultural, legal, and ethical challenges to manage. If a computer makes an error, who is responsible? Who is liable? We don’t always understand why AI makes certain decisions or what biases are in the data. We don’t know whether the use of AI will deskill the workforce. We won’t get an opportunity to resolve all these issues before having to manage AI.
Where should leaders start?
Healthcare leaders should focus first on analytic and backoffice automation opportunities. Electronic medical records, and other systems in the hospital produce more data than humans can use and understand. AI techniques will make sense of the data faster and in new ways. These insights will help to improve decision-making and the redesign of complex operations. Medical advances will continue, and doctors will introduce them into the organization. In those cases health leaders will have to watch for legal, ethical, and labor issues.
Technology always enters healthcare a little slower than other industries – few industries are as complex and high risk. Despite this, AI will move very fast and leaders need to pay attention.
For more on AI, see part 2.
Click here for an article on how hospitals are using AI to tackle care variation.
Photo credit: https://pixabay.com/photo-3706563/
Artificial Intelligence (AI) is a sudden increase in computing ability brought about by recent developments in cheap computer processing power and the collection of vast amounts of digital data. This has the potential to affect almost every aspect of healthcare – clinical, operations, and strategy. Healthcare leaders can think about AI in two ways: the ability to automate more processes, and to gain new insights to inform decisions. We are still at the beginning, but movement here is likely to be very fast owing to the nature of the technology. This will create many opportunities and challenges.