“No app or algorithm will ever be able to do what a GP does”, and “much of what a GP does is based on a trusting relationship between a patient and a doctor, and research has shown that GPs have a ‘gut feeling’ when they just know something is wrong with a patient”, declared vice chairman of the U.K. Royal College of General Practitioners (RCGP) Martin Marshall rigidly (Kahn, J, 2018). Marshall made his statement responding to a claim made by artificial intelligence (AI) developer Babylon Healthcare Services that Babylon’s AI outperformed human General Practitioners (GPs), scoring a very decent 81% on official RCGP qualification exam questions versus a human average of 72% (Razzaki et al., 2018). We can and should interpret Marshall’s audacious response in at least two ways. At first, of course it is true that a smartphone app will never be able to perform certain physical tests or treatments without the use of additional hardware and indeed, the trusting relationship between patient and GP is quite different from the relationship that patients have with their smartphone apps. However, secondly, Marshall’s statement “no app […] will ever be able” is a logical fallacy akin to Charles Chaplin’s assertion, distinguishing between live theatre performances and cinema. In 1916 Charlie uttered that “the cinema is little more than a fad. It’s canned drama. What audiences really want to see is flesh and blood on the stage” (Cerf & Navasky, 1998). Neither Charlie nor Martin appeared to be able to envisage and appreciate the full potential of new opportunities granted by the invention of a new technology. Marshall’s colleague and chair of the RCGP Helen Stokes-Lampard fortunately was more prudent in her response, declaring that “we need to be cautious about the claims being made at the moment. It’s really exciting to see AI evolving and there will be great applications as its gets better, but right now we don’t have independently-verified, peer-reviewed research – the scientific standard for accepting innovation and moving forward” (Wnuk, P, 2018). Stokes-Lampard’s demand for scientific peer-reviewed research makes perfect sense, as the field of medicine is of course rooted in science and the quality of healthcare technology must always be undisputed. Nonetheless, the fact that Babylon Healthcare Services has managed to develop an AI that appears to outperform human GPs in at least some conditions, is an achievement that has many potential benefits across the globe.
Babylon’s new AI system is made available as a voice-controlled skill on Amazon Alexa, while the previous version was only available as a regular smartphone app with a Conversational User Interface (CUI). These types of AI-powered Natural Language Processing (NLP) applications are commonly referred to as chatbots. Users are able to operate these systems by using natural language, rather than clicking buttons. By asking the right questions in the right order, the Babylon AI is able to perform triage and even generates a diagnose and offers a number of possible scenarios, combined with a calculated assessment of each one being correct (Copestake, 2018). Additionally, the chatbot is able to instantly connect the patient with a human GP by means of a video-call, either in case this is deemed necessary by the patient or as a direct result of the triage process. Consequently, Babylon CEO Ali Parsa disagrees with the notion that his chatbot service could replace GPs: “We are very aware that an AI on its own cannot look after a patient and that is why we complement it with physicians. It’s never going to replace a doctor” (Copestake, 2018). Even if the system is incapable of replacing GPs altogether, it still could be very useful in a lot of circumstances where access to healthcare is limited, for example in developing countries such as Rwanda. Whereas patients in the U.K. have access to no less than 2.8 GPs per 1,000 people, patients in Rwanda only have access to 0.0064 GPs per 1,000 people (Copestake, 2018). Having said that, patients and GPs in the U.K. too could benefit from Babylon’s chatbot. By using the structured medical data gathered and analysed by the bot, the GP and patient can use the limited time available during a regular doctor’s visit in a more efficient way. Additionally, it is quite likely that the service will altogether reduce the total number of GP visits, since the information and recommendations provided by the app could often suffice in non-threatening situations.