Why Patients Trust Robots and the Days of Talking to Your Doctor Are Over

Working efficiently in complex systems and delivering complex projects keeps me busy. 

Curiosity about how humans interact with each other in systems keeps me motivated. Ultimately leading me to become an executive coach trained by the renowned Tavistock Insitute in the UK. If you’d like to know more do contact me, in addition, more information will be posted soon on our stunning new website.

Curiosity on how best to increase efficiencies and free up resources has led to a fascination with technology and so I recently visited the “AI more than human”  exhibition at the Barbican in London.  

Focus on both areas and technological advances have led to a fascination with how humans interact with AI.

Read on to find out why patients trust robots and why I am starting to doubt the truth of the following statement  “AI will never replace doctors because patients will always want the human touch”

A recent article I read by a healthcare professional in Switzerland quoted him as saying:

“AI will never replace doctors, because, patients will always want the human touch”. 

I’ve thought about this statement a lot.   It is clearly false. At least the link – AI will never replace doctors because patients want a human touch. The healthcare patients receive is increasingly decided by payer considerations and not by what the patient wants. 

The thinking of this particular doctor was that AI will support doctors, by managing the paperwork and generally increasing the efficiencies of the process. This will then free the doctor up to have meaningful conversations. It does sound good. But he is, I believe, wrong.

Because the system is already penalizing doctors for talking to their patients.

The fact is, that payers are already putting pressure on doctors to minimize the time they spend talking to patients. And the time a doctor does spend talking to a patient is reimbursed at a much lower rate than time spent taking blood, doing an EKG, or doing any other examination. 

In addition, research also shows increasing physician shortages in primary and specialty care worldwide. The predictions for face to face access to human physicians in 30 years are not brilliant. In fact, already now they are not brilliant. 

This means that if AI and automated processes help to speed up the health-care provided to patients, it is likely that additional time that the doctor now has available, will be invested in seeing more patients. Not in spending more time talking to the patients that she already has. 

But perhaps it doesn’t matter. What if we do need a human touch, but we don’t necessarily need humans to deliver it?

We all know that experiencing a pleasant human touch when seeing a physician is not a given

Two examples:

A friend recently visited the doctor because of an allergic reaction.
The doctor said “hello, have you googled it? What do you think it is?”

Another friend scheduled an appointment and discussed her symptoms and was given a prescription. As she was leaving, she remembered another symptom, she had forgotten to mention.  She was about to share this, when the doctor said  “No, we don’t have time, you will need to make a new appointment”

My point is not that all physicians are disinterested in their patients. I know many spectacular physicians.  But as an ex-clinician, I remember the pressure doctors are under to perform consistently and empathetically in an economic environment that is ever less conducive to them performing at their best. 
As a clinician working under pressure it is increasingly hard to maintain a human touch and interest in your patients.

So now contrast these scenarios:

Scenario 1: You go to your GP. You wait for 1 hour before they can see you. Finally, you see a  very stressed doctor, who types notes while talking to you and then hurries you on your way with a prescription. You are not sure you made eye-contact the whole time. You walk out and realize you forgot something. You try to go back. The doctor has a new patient. You leave feeling irritated. 

Scenario 2: You go to your GP. You are shown in immediately. Your doctor is a humanoid robot. She is warm, welcoming, friendly and engaged. She is not pressed for time. She looks you in the eye as you tell her your symptoms, she asks you many questions, lets you finish your sentences and mimics your facial expression. She doesn’t hurry you out the door. 

And ask yourself this. If the outcome for your health is equal, which interaction would you prefer?

You might say you’d still prefer to interact with a human. And if that human is focused on you and in a good space, I would agree. I would too.

But if my options are tired, grumpy, hurried human, or exquisitely programmed AI, I’d prefer the AI. 

A series of experiments done by neuroscientist Julia Mossbridge Ph.D. is interesting in this context. 

Julia is many things to many people. A visiting scholar at Northwestern University, a fellow at the Institute of Noetic Sciences, the science director at Focus@Will Labs, and an associate professor of integral and transpersonal psychology at the California Institute of Integral Studies. Above all Julia has an extremely inquisitive mind and is a joy to talk to. I will be sharing an interview I did with her in a future newsletter.

I met Julia at the Intelligent Health AI 2018 Meeting in Basel. she presented her experiments with robot Sophia. Sophia, a humanoid robot, led volunteers through meditation. Their response to the meditation was monitored with heart rate monitors. They also received questions before and after the interaction.

Watching the videos I was struck by the quality of the human-robot interactions. And by what the volunteers reported they had felt during the interaction.

One volunteer said “I felt much better doing this with a robot than with a human, because, I knew the robot was not judging me”

Despite Sophia’s stilted communication style and missing lower body it seemed easy for volunteers to forget that Sophia is not a human being. 

Another volunteer, with a history of depression, and a very sad expression asked Sophia,  “why do you look so sad?” showing how effective the robot’s facial expression mirroring programming is.  

All the volunteers said, and I paraphrase here, that they felt an increase in their feeling of love for all of creation after interacting with Sophia. If you want more information I can put you in contact with Julia or point you in the direction of her papers.

Julia’s experiment compared readouts from volunteers interacting with Sophia, with a two-dimensional AI on a screen, or with a blank screen and the same programming.

While both the humanoid robot and the two-dimensional figure scored well with the volunteers, the disembodied voice did not have the same effect in generating well-being.

The recent Barbican exhibition gave me the opportunity to observe human/robot interactions for myself.

I interpret Julia’s experiments as showing that if our brains interact with a humanoid image that mirrors our expressions, this is sufficient to trigger feelings of wellbeing. My video at the Barbican indicates that the interaction human/robot looks very similar to a human/pet dog. 
We know our brains evolved very slowly and that the speed with which our environment has changed in the past 500 years has been dramatic. AI has not been around for a very long time. Our brains change slowly. There is no precedent for them to differentiate between robot and human.  So if our brains haven’t been able to keep up, this raises an interesting hypothesis:

Perhaps our brains cannot tell the difference, at a subconscious level, between a humanoid robot and a human. 

Does that mean, in some situations, that you don’t need a human to provide a human touch? 

We like to believe the answer is, that humans are irreplaceable because it makes us secure that we are unique. However, doesn’t it seem much more likely, that our brains rate an emotionally satisfying encounter with a human or a validating experience with a humanoid AI in the same positive way? 

If that were true, then we would not need humans, in every instance, to provide the human touch.

At first glance, this might seem like bad news.  Especially, as most of us, with the exception of Julia and her peers, have thought of the benefits of  AI  versus human as higher accuracy, higher efficiencies, and lower costs.

However, as said above, in the future, we will lack sufficient HCPs both in the developed world, due to the demographic developments, and in many other countries, due to the sheer size of the geographies that need to be serviced and the populations. 

We already face challenges in our healthcare systems to provide qualified support in mental health care units, nursing homes, and many other facilities. 

So perhaps this is good news?

If AI could provide efficiency, accuracy and evoke the same or a comparable response in a patient that a loving human interaction evokes, that might initially seem scary, but at a second glance, it could be very beneficial indeed. 

Experience the Sophia and LOVING AI Project – Meditation Session here on YouTube.

I would love to know your thoughts.

– If you could be treated by a robot or a human. The health outcomes are identical. The robot is well-programmed and comes at 50% of the price of the human and you had the option of having second level human support. Would you go with robot first human after? 

– Do you think AI will free up doctors to talk to patients? Or that more AI will increase speed and give doctors less time to talk to patients?

As ever, If you need help with a project, for example with your medical affairs or medical information strategy, to implement a new IT system, or want to discuss coaching, please contact me to see how I might be able to help.