Dr. Chatbot or AI meets the Botniks

I recently read an article about the National Health Service in England beginning a collaboration with a digital media company to build a Chatbot that can interface with patients.

One of the biggest issues in the British Healthcare System is how long a patient waits to see a doctor and in parallel with that wait, how much of their time doctors spend diagnosing and dealing with relatively simple repetitive issues and basic information transfer.

With artificial intelligence and a chat bot that can take care of doing the routine and mundane tasks of a first-line general practitioner, the physician is freed up to concentrate on the more challenging medical issues her patients face.  Welcome Dr. Chatbot.

a version of a Dr. Chatbot circa 2012

a version of a Dr./Chatbot circa 2012.

Here are a few of issues this quasi medical newcomer must address:

  1. Human like – Studies have shown a higher utilization rate of the tools when there is a more human-centric response. Terminals at a human’s height are most effective. Think of the banking industry deployment of digital assistants in branch.
  2. User Experience – combining big data synthesis and human comfort to hone in quickly on the real problem underlying how the patient presents.
  3. Compliance – it is a well known fact that clinical trials have a high, 50% or higher, fail rate because of failure to “follow as directed.”  This leads to the need for human interaction and a true human intervener to ask the questions listen to your concerns and yes, touch your hand to improve compliance.

The Hawthorne effect is a well documented phenomenon of improved results because the subject knows they are being watched. There is a new startup company called Hawthorne Effect looking to do just that. They are providing  human interveners with the hopes of providing better and more accurate results in clinical trials.

The premise is a sound one in that trusted human professionals go out to patients who are in need of real time real personal interaction  and focused follow up. The patient  may need a reason to see the trial through completion or to “take as directed” the actions or medications to improve efficacy as well as accuracy.

The potential for widespread health breakthroughs with the Dr. Chatbot concept comes from the underlying technologies of machine learning and artificial intelligence.  Newer studies are showing that there are predictors of disease that show up statistically before any one human doctor is likely to reach the diagnosis. The ability to provide an alert to a physician and do the high end mathematical permutations that lead to better and earlier interventions holds tremendous potential. But it will be the user/patient experience that determines compliance. I hope the tech people creating these machine driven interfaces pay particular attention to the human experience (or user experience -as the new discipline is called). Jeff Pulver has called for a move away from the term Chatbot to the more hip, creative and human like Botnik.

Indeed, we need to think of them as Jeff does, ““Botniks are creative, artistic, & rebellious just like the Beatniks” The time is right. There is a shift in thinking from what might work to how elegantly we can begin to deploy the tools. We have entered an era of intense focus on AI, machine learning and robotics and the best winners in this emerging category will be those that enhance the human experience.

Full disclosure: I am a cheerleader for and investor in Hawthorne Effect. I am betting that the human creative can direct the machine and not the other way around.