2017 Trends: Artificial Intelligence Acceleration
Not to get all sci-fi on you, but 2017 is the year that smart machines go mainstream. That’s right: technologies are starting to learn, and that’s a good thing. But if you’re a technology or services company in healthcare, understanding your place in this world is complicated.
What’s the Deal with AI?
Artificial intelligence (AI) is actually just the logical continuation of a series of advancements in technological capabilities. IBM has a useful paradigm when thinking about the history of computing:
- The Tabulating Era (1900s – 1940s): A time when we relied on single-purpose mechanical systems to input and store data.
- The Programming Era (1950s – 2000s): Embodied the shift from mechanical tabulators to electronic systems and programmable computers.
- The Cognitive Era (2011 – ): Our current era where the potential of computer-generated decision-making is becoming real.
So, we’re in the Cognitive Era — but what does that really mean? And what is AI, after all? Well, it’s not just one thing. A useful Wall Street Journal article recently defined it as encompassing “the techniques used to teach computers how to learn, reason, perceive, infer, communicate, and make decisions like humans do.”
Since AI is an exceptionally broad category, some believe it’s not a terribly useful moniker to articulate the specifics of a solution. And some in healthcare are questioning whether it should be used at all, given its tendency to evoke images like this one:
But before we make a judgment call on that, let’s explore AI’s subcategories and adjacent spaces.
How About Machine Learning?
Machine learning is a fast-growing and powerful discipline within AI. It’s an automated approach to analytics that enables computers to draw conclusions from data, all while continually learning and improving. Machine learning often overlaps with another sub-field of AI, natural language processing. This takes text or speech, “understands” the intent of a request, and delivers a response or data set. (Think: Siri.)
Said another way, machine learning is data and analytics on steroids, and incredibly exciting for healthcare — especially when you consider the power of its underlying components such as deep learning and neural networking.
In fact, it’s already gaining tremendous momentum in our industry. Massachusetts General Hospital is one promising example, as they use NVIDIA’s AI server and deep learning algorithms to traverse 10 billion medical images and find anomalies on CT scans, MRIs, and other medical images. (Read more here.)
So, That’s the Same as Cognitive Computing?
Well… it’s complicated. Think of it like this. If machine learning is the discipline, then cognitive computing is the system. Lynne Parker of the National Science Foundation said it best: cognitive systems make use of machine learning techniques, and are “often a complete architecture of multiple AI subsystems that work together.”
There are even more examples of this in healthcare. Take a look at Intermountain Healthcare’s use of CognitiveScale to provide personalized recommendations to adolescents managing Type 1 Diabetes; as well as HCA’s use of Digital Reasoning to analyze and understand massive quantities of clinical data from EHRs and other sources.
In short, healthcare must leverage unstructured data in order to maximize data-informed practice, and cognitive computing is the only way to understand mass amounts of it.
Industry Challenges and Opportunities
The vision is nothing short of inspirational. Early detection of cancers before they metastasize. Unlimited scaling of healthcare interventions and coaching programs. Real-time knowledge of the efficacy of drugs. I could wax poetic for hours here, but I’m running out of word count!
Yet we do have some guardrails. First off, AI only works if you have ample, reliable, clean data. In the lovely words of Heidi Maher of the Compliance, Governance, and Oversight Counsel: “Like children, artificial intelligence needs proper parenting to achieve its full potential, and proper parenting starts with a healthy diet — of good data.”
Further, AI should never be seen as a replacement for clinical care from doctors and nurses; instead, it should augment, inform, and scale that care.
What it Means for Communications and Marketing
A common question ReviveHealth receives from clients in this arena is how to get out from under IBM Watson’s shadow. But realistically, you should think about IBM as your best friend: they’ve done all the heavy lifting, thanks to their brand and marketing heft, of educating the market. And they only represent one sub-field of AI, so there’s great potential to establish your position in the broader market.
The onus is on you to determine how you want to — and can — interact with this emerging field. So here are three takeaways and next steps for communications and marketing leaders:
- Get with your product team to understand whether they’re building any of these capabilities into upcoming solutions, and establish a clear marketing roadmap to build up to and celebrate those advancements.
- Be cognizant of, but not myopically focused on, IBM Watson. Determine how your solution is differentiated from it, and establish a clear strategy of how and when you’ll point to IBM as a reference point.
- If you have a true AI story to tell, begin thoughtfully infusing messages of scalability, intelligence, and iterative learning — and move fast.
Regardless of what you want to call this industry transformation — AI, cognitive computing, machine learning, or none of the above — the impact is going to be massive. I believe 2017 is the year of real liftoff in healthcare.
But I beg of you: be responsible. Please, don’t turn AI into what has happened with population health; please, don’t let it become just another empty term made meaningless by redundancy and forgotten promises.
Because it’s just too momentous, too important. And because only you, healthcare marketing leaders, can prevent the sea of sameness.
Looking for More Resources?
Weber Shandwick just released their research findings on consumers’ opinions and knowledge of Artificial Intelligence, augmented with how well marketers are aligned with those perspectives. See the results.