Doing Simple Better

In Forbes, Dan Gingiss offers this powerful piece of advice for improving customer experience: do simple better.  More precisely, he says customer experience should be:

• Simple instead of complicated
• Fast instead of slow
• Fun instead of boring

Boy, if there is any industry that should take this advice to heart, it is healthcare. 

Mr. Gingiss gives credit for the advice to Chicago Cubs coach Joe Madden, who urges his players to do the simple better -- make the routine plays, you know, routinely.  It's like the advice the manager in Bull Durham gave his young, under-performing minor league team:

Mr. Gingiss draws the analogy: "In business, this equates to making every interaction with the customer easier."   He compares, for example, Google's home page to Yahoo's; the former is spare, almost sparse, while the latter is needlessly busy.  Remind me, which one dominates search?

"Simple" is not a word most people would usually use to describe their interactions with the healthcare system.  Nor is "fast" or "fun." 

Look at your health insurance policy, or a healthcare bill, or the warning labels on prescriptions.   Any of those simple?  Is waiting for your physician or test or service fast, much less fun?  You're probably got a patient portal, or two or three, plus one for your health insurance, but do any of these really make things easier for you?

No, in healthcare we make the simple complicated, we make fast slow, and we make fun at best boring and worst scary. 

It's useful to remember that for most of the time there has been healthcare, care was more about comfort than cure.  Doctors and other healers didn't have much in their bag of medical tricks, but what they could be counted on was to comfort sick people -- maybe give them hope, sometimes even help them get better, but always be a source of comfort.

That bag of tricks is pretty big now.  We have some solid understanding of what causes many -- not all -- diseases and disorders, and numerous treatment options to deal with them.  In fact, we expect our physicians to always be able to do something for us: give us a pill, perform some surgery, get us into a clinical trial when all hope seems lost. 

But we've lost simple along the way.

Our doctors are rushed each time we see them.  Nurses, that last bastion of comfort, are usually in short supply, and often replaced by a confusing array of other medical personnel -- LPNs, medical assistants, medical technicians, and so on.  We don't have one diagnosis, we have many. We don't have one doctor, we have several.  We don't take one prescription, we take multiple. 

We're getting plenty of care, but comfort often ends up being short-changed.  The complicated is driving out the simple.

Health is complicated; there are no shortcuts to it.  Healthcare is complicated; figuring out exactly what is wrong with someone is hard, as is tailoring the treatment to them.   Complicated now comes with the territory, but we're letting it define healthcare.

I still choose to believe that most people involved in healthcare want to help people.  I suspect most of them still believe that what they are doing is helping people, and, in many cases, it may be.  But I don't think that enough of them are heeding Mr. Gingiss's advice; they're not working to make things simple, or at least simpler.

Mr. Gingiss warns: "Simplicity is a basic tenet of customer experience, but it is often overlooked in favor of a company's outdated rules or procedures."  Indeed, too much of healthcare is about following the rules and regulations.  Too much of healthcare is making sure the various parties get their money.  Too much of it is about the convenience of the people working in healthcare.  Too much of it is about trying to avoid getting sued.

At every touchpoint, at every interaction, we should be asking: is what I'm doing making it simpler for the patient/consumer?  Is it happening as fast as they are hoping?  How can I make it more fun for them -- or, at least, how can I make it more comforting for them?

These are not the questions that we are asking.  These are not the goals we are setting. These are not even things we are measuring.  As a result, too much -- way too much -- of healthcare ends up being about what we're doing to people, not what we're doing for them. 

Do simple better.  Help them feel better.  Comfort them. 

Mr. Gingiss concludes:

Take a hard look at your company’s experiences and customer touchpoints – both offline and online. There are so many opportunities to Do Simple Better and because they are simple by nature, they are also usually inexpensive to execute. It’s the little things that matter in customer experience, and a lot of little things can go a long way to differentiating your company’s experience from that of a competitor.

We can do better.  We have to do better.  We may or may not be able to make the healthcare experience simple, but no one can doubt that we can make it simpler.  We can also try to make it faster, and we should try to make it more fun. 

But will we?  What did you do about it today, and what can you do about it tomorrow?

I'll close with another one of Albert Einstein's great quotes:

Help Wanted: Tech Big 5, Please Apply

In such a technology-dominated era, I keep thinking about why it is that there are no true healthcare technology giants, companies that other industries are watching closely, worried they will come into their space -- the way healthcare is thinking about tech companies.

In some ways, it is the best of times for big tech companies.  Tech companies are the five biggest companies in the world by market cap.   Jeff Bezos is now the richest man in the world.  Even people in third world countries increasingly have access to smartphones, not to mention Facebook, online shopping, and Alexa.

In other ways, it is the worst of times for them.  We worry about our personal data;  we lament the decline of retail shopping; we realize that many of us may suffer from "smartphone addiction."

But few of us are anxious to go back to a 1990's era of technology.

Amanda Lotz, a professor at University of Michigan, recently pointed out that the Big Five tech companies -- Alphabet, Amazon, Apple, Facebook, and Microsoft -- emerged from and still dominate largely distinct businesses.  At the risk of oversimplifying:

• Alphabet seeks to make information easier to find and understood; Google dominates online searches.
• Amazon seeks to make it easy to find and purchase products; it dominates online shopping.
• Apple seeks to make beautiful, easy-to-use computing hardware, with proprietary software running it; it dominates the smartphone market (especially in terms of profits). 
• Facebook seeks "to bring the world closer" through social networks; it dominates social media.
• Microsoft seeks to be our operating system (OS); it still dominates the desktop/laptop OS market.

Source: James Schad, The Relentless Rise of the Tech Giants

To be sure, the lines are blurring.  Alphabet has the largest mobile OS; Alphabet, Amazon, and Microsoft all make devices; Microsoft has the largest business social network; Apple and Amazon have large entertainment businesses; they all, it seems, want to be leaders in cloud computing and AI.

And each of them is rumored to be exploring moves into healthcare.

Let's think about a world in which each focused on healthcare as its core business:

Alphabet
Healthcare has lots of information.  It supposedly has 30% of the world's stored data.  There are several hundred thousand research studies annually.   Unfortunately, much of the data is siloed, many of the studies are behind paywalls and/or aimed at only the most hyper-specialized professionals.  Most healthcare language is jargon.  No one can absorb, understand, or make good use of all this information.

This cries out for a company that can find, make sense of, and communicate the appropriate information to the necessary audience(s).  It makes searching the world's libraries look like child's play.

This cries out for Google.

Amazon
Americans are a nation of consumers, and we have the spending to prove it.  We spend more on healthcare than anything else, yet do a worse job buying it than we do anything else -- things are too inexpensive, prices are wildly inconsistent, and many things that we buy are of little or even negative value.   

We don't really need to find the cheapest HDTV or the cutest sweater, but we do need to find the right healthcare.  We need help comparing healthcare prices, clinician quality, and value of treatments.  We need help in purchasing healthcare services in ways that are fast and frictionless.  We want to go to a healthcare retail location that someone has fully reimagined for us, as Amazon is doing for bookstores or convenience stores.

This cries out for Amazon.

Apple
Healthcare has lots of technology.  Most of it is expensive technology, like Defense Department-level expensive ($30 screws?  Try $30 aspirins).  Unfortunately, little of healthcare technology delights its users, whether those users are clinicians or patients.  Interfaces are often epically bad, data doesn't flow yet often doesn't stay private, and we suspect that an industry which still settles for faxes is making technology compromises elsewhere too.

We need healthcare devices that are easy to use.  We need healthcare devices that we long to have.  We need healthcare devices that are part of an ecosystem that fits together and that communicates well, while keeping our data private.   We need healthcare devices that we can count on not to fail us when we need them most.

This cries out for Apple.

Facebook
Being healthy is hard.  It requires developing and maintaining the right habits.  Being sick is worse.  It is a scary time, full of uncertainty and fear.  For both times, we need help, and some of that help must come from our social networks.

We need to find the people in our lives who will inform and support our good health habits.  We need to find the people in our lives who have, or have had, similar health challenges, who can tell us what to expect, what to look for, and where to look for it.  And we need to be able to provide the same kinds of support for the people in our lives who need it from us. 

This cries out for Facebook.

Microsoft
Let's face it: our lives are not well designed to do the kinds of things that will help us keep healthy.  Our lives are built for too much sitting, too much eating, too much stress.  You could probably design a lifestyle OS aimed at producing worse health results, but it wouldn't be easy.

As Steve Downs called out for his great series, we need to build health into the OS of our lives.  We need to make better lifestyle choices, and be more mindful about the impacts on our health.  Some of this can be through technology, some of can be through how we structure our daily lives, and some of it has to be through how we view our health.

Through his foundation, Bill Gates is trying to impact health problems in the developing world; Bill, we have a crisis right here.  We need a new health OS.

This cries out for Microsoft.

It doesn't have to be all of these, or even any of these.  In some ways, it might be better if it were new companies, who are staking their bets that they can become the new tech giants by taking on healthcare.

Either way, I want to see a healthcare system whose tech is envied and emulated by other industries.

Animal Farm Meets Health Care

In George Orwell's classic Animal Farm, the animals revolt against their human masters, and establish a classless society with the inspiring principle, "All animals are equal."  As events play out, their society devolves into a dictatorship with a ruling elite, and the principle becomes "All animals are equal, but some are more equal than others."

This, surprisingly, makes me think of health care. 

I am old enough to remember when maternity coverage was at best only very limited even in employer group health plans.  It took the Pregnancy Discrimination Act (1978) to require them to treat maternity the same as any "illness," and, even then, individuals plans often did not include it until ACA required it.  Similarly, coverage for mental health was typically skimpy until the Mental Health Parity Act (2008) required parity.

Preventive services were usually only available for (the small percentage of) people enrolled in HMOs, until network-based managed care plans grew more widespread in the 1990's.  The same happened with prescription drug coverage, which used to only be available to the minority of people with "major medical" coverage.

It took the Affordable Care Act to standardize what "essential benefits" should be included in health plans, and even then they are almost entirely medical benefits (and are still under attack, such as with "short term" health plans or Idaho's recent attempt to sell low cost plans). 

But, for the most part, medical services are generally covered by (most) health insurance plans.  For services like dental, vision, or hearing, not so much.   Evidently, some services are more equal than others.

We've managed to push our rate of people without health insurance to around 11%, but it's more than double that for dental insurance, and worse yet for vision coverage.  For seniors, the figures are significantly worse. 

The real question should be, why do we have separate coverages for services like dental or vision, especially when many lack them?

This matters.  According to NCHS, 14% of Americas report hearing trouble, 9% vision trouble -- and 7% have no natural teeth left (25% for those over 75).  There is a well documented link between oral health and our overall health, yet a study found that dental care had the highest financial barriers to care, compared to other health services.

Percentages of National Health Interview Survey respondents who did not get selected health care services they needed in the past 12 months because of cost, by age group, 2014.  Vujicic, et. alia.   

A third of millennials -- whose teeth one would expect to be good -- are reluctant to smile due to their teeth; 28% say it impacts their ability to get a job.  Poor oral health is most pervasive among low income people, yet only 17 states have Medicaid programs with comprehensive adult dental benefits, and the trend is to require that able-bodied people get jobs in order to qualify for Medicaid coverage.

Kind of a Catch-22.

We claim to support preventive services -- to the point ACA mandates coverage at no cost -- but that only applies to services that are medical in nature, performed by physicians.  If you want to get your eyes or your teeth checked, maybe you'll have some coverage for it, maybe it will even be covered at 100%, but, then again, maybe not to either.

Austin Frakt laments: "It’s an accident of history that oral care has been divided from care for the rest of our bodies. But it seems less of an accident that the current system hurts those who need it most."  Nicholas Kristof recounts that historical accident:

It’s virtually an accident of history that dental care isn’t considered part of medical care. The medieval barber-surgeon used to attend to all the human ailments that required a knife: bloodletting, tooth extraction, shaving. In the 1840s in the United States, the heirs to the tradition wanted to become professionals; they didn’t want to keep wandering from town to town selling their services. They asked physicians at the Medical College at the University of Maryland if they would include dentistry in the medical coursework, but the physicians refused. Soon after the dentists opened a separate dental school nearby.

And keep in mind, those weren't post-Flexner educated physicians.

If you break a bone, you'll see a doctor; if you break a tooth, you'll see a dentist.  If you have problems with your throat,you'll see a doctor; if you have problems with your gums, you'll see a dentist.  If you want to correct your vision with glasses, you'll see a optometrist; if you want to correct it with Lasik, you'll see a physician.

Think about the alphabet soup of "doctors" we have: M.D., D.O.s, D.C., Ph.D., PharmD, O.D., and D.P.M., and I've probably missed some. 

Specialization is understandable, as most physicians end up doing, but I have to wonder why some types of specialization start at the beginning of training, rather than after the basic medical training (see my previous article on balkanized medical education).

We accept all this because, well, that's the way it always has been.  That doesn't mean it makes sense, or that it is best for our health.

For example, the microbiome is a hot area of medical research and is starting to be an important option for treatments.  What if, say, microbiologists claimed only they could understand it and should be the only practitioners allowed to treat microbiome-related issues?  It would force us to guess when to see whom for what, but that is essentially what we've done with many other aspects of our health.

We each only have one body.  Although some health issues are fairly specific, we are increasingly realizing that many are systems issues involving multiple parts of the body.  It's time to stop drawing artificial distinctions between what care we get, who gives it to us, and how those professionals get trained. 

Health is not equal to health care.  Health care should not be limited to medical care.  We need to get past "historical accidents" and focus on what is best for our health, and our care.

Unless you actually do believe that all health services, and all health care professionals, are equal, but some are more equal than others.

Healthcare's Death Star Thinking

I missed it when it first came out, but a providential tweet from the always perceptive Steve Downs tipped me to a most interesting article from Jennifer Pahlka with the wonderful title “Death Star Thinking and Government Reform.”

The article is not directly related to healthcare, although it does include healthcare examples, but Ms. Pahlka’s central point very much applies to most efforts to reform healthcare:

The need to believe that a Death Star-style solution is at hand — that we have analyzed the plans and found the single point of failure — runs deep in our culture.

After all, who among us does not want to be the Luke Skywalker who blows up healthcare’s dysfunctional Death Star? 

Of course, in reality, figuring out exactly who the evil Empire of healthcare is or what they’re using as a Death Star these days is not so easy. Nor is finding the single point of failure that we can exploit to magically make everything better.

Some see the health insurers as the Empire (or, certainly, their supposed agents the PBMs — the Darth Vaders of healthcare?). Others blame the ever-consolidating health systems, which are gobbling up fellow hospitals and any available physician practices. Pharmaceutical companies are also a popular target. 

We can’t quite decide if physicians are the Imperial storm troopers who oppress us or the persecuted Jedi who may yet save us (although many — including a lot of physicians — feel that the AMA is part of the Empire). Nor can we decide if employers are the Imperial tax collectors who have all too passively allowed healthcare costs to steal our wages, or the plucky Hans Solos who are doing their best to fight the Empire. 

Proposed magic bullets to fix supposed single points of failure in healthcare abound. Remove the tax preference for employment-based coverage. Give us more “skin in the game” via increased cost-sharing. Increase the number of physicians, or at least reduce the percent who are specialists. Increase integration between health professionals/facilities. Get more of us covered (or maybe not).

Improve interoperability. Make the healthcare system more digital. Use artificial intelligence (AI) and Big Data to figure out what we humans haven’t been able to. Nanotechnology or robots, even nanorobots, will save the day. Develop even more new drugs, use 3D printing, take advantage of gene splicing/editing. Exploit the microbiome’s power. 

Maybe we’re looking in the wrong places and we should be focusing on social determinants of health (SDOH). 

Or we could all just eat better and walk more. 

A sobering new study from Ashish Jha and colleagues reminds us, though, that there probably is no such single point of failure. As other studies have attempted to do, it compares the U.S. healthcare system to other developed countries. It concludes that, aside from our high level of spending (caused by high prices and administrative costs), we’re not as bad or as unusual as many think. 

It’s not that any country has fixed the single point of failure as it is that every country is also struggling to find it. Harvard Medical School professor Bruce Landon told The New York Times: 

I don’t think there’s any of these countries where if you went and talked to them individually, they wouldn’t say they’re having a health care cost crisis. They’re all struggling with paying for new technology and the cost of the system.

We’re all engaging in some Death Star thinking.

Whatever the single point of failure might be, many think we need a new hero to fix it, be that Amazon, Google, Apple, or people/companies we haven’t heard of or aren’t thinking about in this space. They’ll come along to rescue us and our health care. For all those people who believe some tech company can ride in to save healthcare, Ms. Pahlka reminds us:

In fact, simply adding technology without understanding the complexity of the bureaucratic processes and how they got to be the way they are can perpetuate the problem, or even make it worse.

She also emphasizes — and this really fits healthcare — “You can’t fix the problem of a system that is based on rigid rules by specifying new rules.” We’re not going to simply legislate or regulate our way out of the mess we’ve allowed our healthcare system to get into. 

Instead, she points out:

There is no “solution,” only continued storytelling that shapes a new narrative and supports a new sensibility. But it is an important, valuable agenda that will get less attention and resources than it should because it doesn’t fit the model of a “solution.”

Fixing healthcare is very much a design issue, one that requires new ways of thinking, better “storytelling,” and involvement of all impacted parties. None of that is easy nor natural for healthcare; as healthcare design guru Dr. Bon Ku said at his recent SXSW presentation: “Healthcare is the black hole of design.” (thanks to Danielle Ralic for the quote):

Dr. Ku is attacking the problem at one of the roots, during medical school, as is Clay Johnston at Dell Medical School and a short-list — too short — of other medical schools. 

A new article in Harvard Business Review by Bhatti and colleagues urges that healthcare use more “human-centered design.” They studied three healthcare innovation centers, identified six common challenges, and made several recommendations for others interested in accelerating human-centered design in healthcare. 

Perhaps their most important statement was a mindset: 

This model isn’t just about getting greater patient feedback during the innovation process. Patients are co-designers, co-developers, and increasingly more responsible for their own and collective health outcomes.

The Empire would hate that. 

If the Star Wars franchise has taught us anything, it may be that there will always be another Death Star. There will always be a would-be Empire. It is never easy to defeat the forces (small “f”!) that seek to keep us from fulfilling our destiny, even if that destiny is better health. But it also reminds us that we should never give up hope.

We didn’t design our way into our current healthcare system, but it is possible — just possible — that we can design our way out of it.

Life in the Cloud

The cloud is big. Amazon (AWS), Google (Cloud Platform), IBM (Cloud) and Microsoft (Azure), among others, have made big bets in the space. Indeed, virtually all of Amazon’s 2017 operating income came from AWS. Alphabet’s former Chairman Eric Schmidt had some important advice for attendees at HIMSS18: "Run to the cloud!"

The thing is, I’m not sure we’re thinking about the cloud quite right yet. As evidence, I want to point to the Blade Shadow PC.

Blade is a France company, and Shadow PC is an app designed specifically for gaming. It offers users a virtual machine, giving them access to a high-end PC that is constantly upgraded.

Gamers always want to have PCs with the fastest processor, best graphics card, and sharpest display. It can get very expensive, and as soon as you’ve upgraded your device, it starts to become outdated. With Shadow, in theory, you’ll always have the latest and best.

Shadow doesn’t even care what device you run it on. Your monthly subscription lets you run it on any device you want (one at a time). You could start a game on your home PC, continue playing on your smartphone while you are taking an Uber to work, then pick up playing it on your work computer (during your breaks, of course). The machine you use is pretty much just needed for the connectivity and the screen.

Shadow has received a lot of media attention. The consensus seems to be that it may not always quite live up to the hype, yet, but it is pretty cool. 

Here’s CNET’s review:

I have to admit, I don’t really care about video games, except as a business and cultural phenomena, but the implication of Shadow PC is not so much for video games as it is for everything else we do on devices. As Blade CEO Asher Kagan told The Next Web: 

If we can do something to prove this is working for the gamer market, it’ll show it can work for anyone. Gamers are very demanding, they’re very sensitive to latency.

The Wall Street Journal’s David Pierce sees it leading to a new future:

Someday soon, what Uber did for cars and Netflix for TV will happen with computers. Rather than buy a suite of gadgets — a PC, phone, Xbox, etc. — you could just access their features when you need them. They won’t need to be distinct, powerful devices with their own hefty allocations of processor and memory. Instead, you’ll have a single virtual computer with all your data and preferences. You’ll reach for a touch screen when you’re on the go, sit down to a larger one with keyboard and mouse when you’re at your desk. Maybe you’ll have a wall-size one, too.

Similarly, Rob Enderle writes in Computerworld: 

Blade Shadow PC has the potential to become a Netflix for PCs but so does Netflix and all the other streaming providers — including Amazon. I expect by 2020 we’ll have a lot of compelling alternatives to a running apps on locally on a PC and that this will drive a trend similar to what happened to Blockbuster and Netflix. The old will give way to the new and we’ll never have to worry about patches, replacing hardware, or even buying apps the way we do now.

Mr. Enderle’s article is headlined “By 2020, we’ll be using Windows in the cloud,” so we’re not just talking about games or apps or websites. We’re talking about operating systems. It takes the idea behind Chromebooks to the next level.

At this point, it might seem that everything will go to the cloud. After all, with cyberthieves getting ever more sophisticated, it almost seems negligent for you to have your own PC or for your company to hosts its own machines. Let the experts worry about hacking and security, about hardware and software upgrades. We only care that things work.

I don’t think it is going to be quite that simple.

I’m old enough to remember mainframe dumb terminals with arcane commands. I’m thus old enough to remember when PCs came along to help wean us from mainframes, when local computing power was something to be prized, not avoided. 

Moving everything back to a centralized place, this time in the cloud, thus leaves me with a little trepidation. 

I don’t think that is what is going to happen because of ubiquitous computing (also known as pervasive computing), and dispersed computing. They sound similar, and they are related, but they are distinct.

The former is often thought of in the Internet-of-Things gold rush, where everything is connected to the Internet, communicating all the time. It is happening already, and it will develop exponentially over the next few years. We’ll know more, about more things, than we ever could have guessed we might have needed.

The latter, though, is an even newer idea (Darpa is spurring it): all those devices aren’t just connected but they are all also computing. You draw your computing power from whatever is handy and appropriate to/necessary for the task(s) at hand. 

You might access a distant cloud server farm for some needs, but your smart clothes for others. If you lose a connection, or if your computing needs shift, you seamlessly pick up another, or add more. Your screen might just be something you see through your AR contact lens and your “keyboard” might just be your hand gestures, or even a direct implant in your brain.

Credit: Darpa

Think of computing power almost like electricity: it is just there, everywhere, and you won’t even always need wires. You use what you need, and you don’t really care where it comes from. 

Think about your computers as, well, you won’t have to think about them at all. As I said in a previous post, 

If you’re aware of your device, that’s the past… We’re going to have to get past our fascination with the latest and greatest devices — a new iPhone! a 4D television! — and let their technology fade into the background. As it should.

Healthcare is very proud about how it is finally adopting (if not quite embracing) computers, and many of its thought-leaders are taking Mr. Schmidt’s advice by starting to move to the cloud as the “next” big thing. That’s good. 

I just would like to see them spending more time thinking about the next next big thing.

Will AI Docs be MDs, or DOs?

There is increasing acceptance that artificial intelligence (A.I.) is going to play a major role in healthcare and in the practice of medicine.

Some see AI as a way to augment human doctors.  Some see it as a way to help patients triage the need to see a human doctor.  Others see it replacing entire specialties (pathology is often cited).  A few even think that, eventually, AI "doctors" could replace humans entirely.

Whatever is going to happen, we need to be thinking about how AI makes its decisions -- and what that might say about our existing system.

AI decision-making has two separate but very much overlapping problems: the "black box" problem and unintentional biases.

The black box problem is that, as AI gets smarter and smarter, we'll lose track of what it is doing.  "Machine learning" refers to the ability of AI to, essentially, learn on its own.  It looks for patterns we might only not have seen, but might not even be able to see.  It may reach conclusions using a logic that is beyond us.

My favorite example of this, as I have previously written about, is Alphabet's Deepmind program AlphaGo Zero.  It learned how to play the fiendishly complex game Go, without humans programming it to play, or even by learning from human games.  It not only mastered the game -- in three days -- but also came up with strategies that left human Go experts agog.

Think about the day when the AI's strategies are not about a game but about treatments for our health.

If we were convinced AI was always making purely objective decisions, we might grow to accept its decisions without question.  After all, most of us don't know how our televisions or smartphones work either; as long as they do what we expect them to, we don't really care how.

The trouble is that we're becoming aware that all-too-human biases can be built into AI.

Programming remains a largely male profession, and those men are usually young, well educated, and from comfortable backgrounds.  It is not a good representation of the world.  Their world views and experiences influence the way they program, and the data that they give to their programs to help them learn.  In most cases, they're not intentionally biasing their creations, but unintentional biases can have the same result.

For example, a recent study from MIT and Stanford on facial recognition AI programs found that they worked very well for faces that were white males; otherwise, not so much.  The researchers discovered that the dataset of one such program was 77% male and 83% white.  The programs were from major technology companies, and certainly weren't intended to be biased, but the AI knew best what it knew most.

This is not the only such example.  As Kriti Sharma points out that the gender stereotyping in having default voice for digital assistants Siri or Alexa be female, but using male names for problem-solving AIs IBM's Watson and Salesforce's Einstein.  ProPublica found that AI used by judges during sentencing to predict the likelihood of future crimes greatly overestimated the likelihood for African-Americans.

The key to both problems may be to increase transparency about what the "black box" is doing.  The Next Web reported that researchers recently "taught" AI to justify its reasoning and point to supporting evidence.   Whether we'll have the time, interest,or expertise to examine these justifications remains to be seen.

Vijay Pande, a general partner at Andreessen Horowitz, isn't so worried, and he specifically points to health care as a reason why:

A.I. is no less transparent than the way in which doctors have always worked — and in many cases it represents an improvement, augmenting what hospitals can do for patients and the entire health care system. After all, the black box in A.I. isn’t a new problem due to new tech: Human intelligence itself is — and always has been — a black box.

We don't really know how physicians make their decisions now, which may account for why physicians' practice patterns are so varied.  Many supposed rational decisions, in healthcare and elsewhere, are based on a variety of factors, most of which we are not consciously aware of and some of which are more instinctive or emotional than intellectual.

Mr. Pande views the so-called black box as a feature, not a bug, of AI, because at least we have a chance of understanding it, unlike the human mind.

Credit: Scott Adams, Dilbert

All of which led me to a thought experiment: as we program healthcare AI, would we want it to be based on allopathic (M.D.) or osteopathic (D.O.) practices?

These are the two major schools of modern medicine.  Their training and licensure have become much more similar over time, but the two remain distinct branches, with separate medical schools and graduate medical education.  Most hospitals are "integrated" but there remain predominately D.O. hospitals.

If you asked either type of physician if the healthcare AI of the future should be based  solely on their own branch, I suspect most would find that acceptable, but not if based solely on the other's.   If you asked if it should be based on both, using all available information, I suspect that would be even more acceptable.

Therein lies the problem: if we don't want our AIs to be either "M.D." or "D.O.," but rather a combination of the best of both, then why don't we want the same of our human doctors?  Why do we still have both?

IBM's CTO Rob High spoke to TechCrunch of the AI work they've done with Sloane Kettering Cancer Center, and admitted the resulting AI has their biases and philosophy.  He says "any system that is going to be used outside of Sloane Kettering needs to carry that same philosophy forward."

Whether it is Sloan Kettering, The Mayo Clinic, or The Cleveland Clinic -- or M.D. versus D.O. --  we should want AI based on as much data as possible.  We don't yet really know what is important, and should not make the same mistakes with silos as we've made before AI.

I don't want my healthcare AI to be either an M.D. or a D.O.  I don't want it to be a physician at all.  I want it to be something new.  If we want the healthcare system of the future to be an improvement over what we have now, we need to stop thinking within our current paradigms.