Last week I went to an event focused on the Amazon Alexa and applications for the management of diabetes that were developed in response to the Alexa Diabetes Challenge. This event, produced by Luminary Labs and sponsored by Merck, was an interesting milepost in the evolving quest to make voice a de facto user interface in the healthcare world – interesting both because there is clear progress and because we are definitely not there yet. As with all things healthcare, nothing happens overnight.
Two years ago I wrote a post on this idea called The Next Voice You Hear May Be Your Own (or Dr. Alexa’s), noting that voice as user-interface in healthcare might well be the next 200,000 year old overnight sensation. Make that 200,002 years and counting. The applications shown at the event were very interesting, but I am not sure they were really giant breakthroughs. Rather they were, for the most part, versions of existing products that had voice as a key user interface, though not even the only one in most cases. In a few of the products voice was nice, but not necessary. Given that Alexa still is not HIPAA compliant (although rumor has it this is fast approaching), there are also still other issues needing attention.
Coincidentally, last week my iPhone served me up this story about a woman who could not figure out where in the hell all of the Amazon purchases were coming from until she realized her parrot was ordering them. And if a parrot can mimic well enough to order from Amazon, imagine what a decently skilled human impersonator could do if they really wanted to. This may well be the full employment act for Ursula from the Little Mermaid if we are not careful.
Security and privacy are essential for this user interface to make it to actual prime time, as it is generally recognized that Minimum Viable Products (MVPs) are not particularly acceptable in healthcare. Yes, Amazon may make it’s way to being a full service pharmacy including delivery, but if Alexa hears Aricept instead of Aciphex, the consumer is going to have more than just a bad taste in their mouth. Though I will note that as I sat there watching the diabetes product demos, I looked up whether you could buy glucose strips at Amazon. Answer: yes. Full service product, here you come.
These issues will ultimately be figured out and we will find voice applications in our everyday healthcare lives before the end of the decade. I just don’t think this is really going to start with consumers. Most think about Alexa and similar products as consumer-facing technologies but I have a feeling that the enterprise will be the first place where the product finds true adoption. Imagine physicians being able to speak to the EMR instead of working all night inputting notes (or hiring scribes – sorry scribes, you are going to be replaced by a robot after all). Voice technologies would be extremely useful in the operating room to get medical information on demand: Alexa, what is the patient’s ejection fraction? Alexa, please show me the last MRI scan. Alexa, what can we bill for that we haven’t thought of yet?
I was, in part, attending the Alexa event to get a better understanding of the investment opportunity in voice. Unfortunately, I came away thinking there isn’t really one, at least not now. Rather, it is going to be a technology that becomes integral to many healthcare products but isn’t one in and of itself. The product is the Echo or the Google home (notably nearly no one in the audience used the latter while many in the audience use the former) or whatever other gizmo is hot at the time. There won’t likely be an Echo for healthcare, as “regular” devices are winning the day even in healthcare. If security and privacy and sensitivity and specificity can be as good as human interaction, we may as well use the devices we can now buy at Whole Foods. No really. My local Whole Foods has a display of Echo devices right next to the organic broccoli. Alexa, can you make broccoli taste like ice cream?
I suspect that the people who make money off of voice as user interface will be design firms that help healthcare companies, often tone deaf about user interface, learn to use voice effectively. Since by and large those are consulting companies, they are generally not considered great venture investing opportunities, but the opportunity to prosper here could be significant for those in the consulting business. I can only imagine how some of our best-known healthcare companies would integrate voice left to their own devices. I’m guessing the most used question would rapidly become: Alexa, what in the hell are you talking about? Or actually, Alexa, I don’t know my PIN code to access my record – if I did I would have looked it up on my damn iPhone. Alexa, where is my damn iPhone?
Most of the applicable voice interfaces are or already have come to the phone so that may be yet another reason why we may as well have our phones implanted in our appendages so they can’t so easily fall in the toilet. Alexa, please dry out my phone. And watch out, Ursula may be down there in the water.
If you are interested in the panel discussion that Luminary Labs hosted about voice in healthcare, you can watch it HERE. Besides me, John Brownstein, Chief Innovation Officer at Boston Childrens Hospital, and Ifeoma Ajunwa, a lawyer, sociologist and Felow at the Berkman Klein Center at Harvard, were on the panel, moderated by Sara Holoubek, CEO of Luminary Labs. FYI, the winner of the Alexa Diabetes Challenge has yet to be announced but the announcement is coming soon. See this space if you are interested. Or hear it if you are at Whole Foods.
Maneesh Juneja says
Lisa, Interesting post as always. When it comes to smart home devices like the Echo, I came across a great article reminding me that whilst the Echo and others are very useful for those living in large multi-room homes in North America, the devices and the voice interfaces are less useful across many parts of the world where homes are much smaller and have fewer rooms than homes in North America. I still believe that from a researcher’s perspective, Amazon may end up with a big repository of longitudinal voice data which might make an attractive cohort for those wanting to see what they could discover in there with respect to digital biomarkers. Additionally, whilst not an immediate value add, assistants with voice interfaces when integrated properly into a smart home may eventually help more seniors ‘age in place’ which not only could save costs, but may even give the user a greater sense of dignity and independence. Looking at some reviews on Amazon, it appears that adult children are buying Echo devices for their parents/grandparents living with Dementia. Overall, I agree with you that we are not quite there yet when it comes to voice interfaces, but I also see this space evolving fairly rapidly, so fingers crossed.
Lisa Suennen says
Hi Mannish, I definitely agree that it has potential for the senior market, particularly those less facile with cell phones. Lisa
As a technologist, my mouth waters at the possibilities of voice-based interfaces and the human interaction patterns. The challenge I have with voice is that it is a lower bandwidth medium than text-based or visual interfaces. So from a day-to-day provider’s standpoint, it may not be as efficient. On the other hand it provides the opportunity to “intonate” meaning and bring life into conversations. There are also several opportunities that can be created particularly in situations where someone’s hands are already in use or if they are taptically challenged..
More than the device, I think the idea of Alexa-as-a-service (that accompanies you wherever you go in whatever form factor you choose) is powerful.
The investment opportunities in voice are as you said, likely to be in companies that integrate voice-based interaction patterns to improve patient outcomes.
I also think there is a good opportunity in motivational interviewing-based virtual coaches to educate patients on lifestyle change, basically a rehash of the virtual avatar paradigm but voice-based. The challenge there is, you have to get it 100% right, even 99.9% will not do.
Lisa Suennen says
Harsh, we are on the same page! Lisa
One of the most pressing issues that is overlooked when discussing emerging healthcare technology and the successes, failures and inability to monetize in the space is human capital and demographics. The US population is growing (326m); aging 10k baby boomer per day; a projected shortfall of 90k physicians; 1m nurses and oh by the way, 75-80% of the population living in urban areas.
While the VC appetite for investing is weighted in software, the innovation is being driven by solutions to support the consumer, save money and drive organizational efficiencies.
Lisa, that said, what is important is that the article is focused on healthcare innovation and not medical treatment. Great article.
Lisa Suennen says
Sherif Khattab says
Thank you for this post. In this world of change, your shared thoughts always mix the excitement, reality, skepticism and more. It makes you think!
Then you inject a little smile into a very serious subject matter. The essence of this post reminds me of two Adobe commercials that always make one laugh no matter how bad the day has been. The “Click Baby Click” & the “BS Detector”. Keep the thoughts flowing and thank you for sharing.
Lisa Suennen says