Chatbots are trending right now but the idea isn't new. How did we get from a simple, basic conversation with machines to efficient symptom diagnosis?
If a patient said “my head hurts”, Eliza would respond, “why do you say your head hurts?”. But Eliza was not a doctor, or even a person. It was a program – 200 lines of experimental code written 50 years ago at MIT. This program was proof that it's possible to simulate human conversation using software.
Eliza was the first chatbot ever created. Chatbots are a new type of user interface which takes advantage of artificial intelligence software that can answer questions, converse, and assist us in a way that mimics human conversation.
The first chatbot simulated a psychotherapist, but it wasn't very human. In 2017, however, you can create a doctor-like chatbot that can pre-diagnose your symptoms before you visit a real doctor.
History of chatbots – 1966 to the early 2000s
Eliza was named after Eliza Doolittle, a working-class character in Pygmalion (a play by G.B. Shaw) who is taught to speak with an upper-class accent. But Eliza the program was taught to simulate a Rogerian psychotherapist.
The software wasn’t nearly as sophisticated as current chatbots, but the 200 lines of code were enough to create the illusion of talking to a psychotherapist. Ultimately, however, Eliza wasn’t capable of passing the Turing Test.
If you can't say whether you're talking to a person or a machine, the machine passes the test. You can talk to Eliza for a minute right here to see why it failed. Apparently the illusion works best when you talk about yourself (which is what you would be doing in a Rogerian psychotherapists office).
The first chatbot to pass the Turing Test was Parry, first implemented in 1972 by psychiatrist Kenneth Colby. Unlike Eliza, Parry imitated a paranoid individual and questioned everything it was told. Parry passed the Turing test in an experiment in which human interrogators questioned him and were unable to tell whether they were talking to a person or a machine.
“The first chatbot to pass the Turing Test was Parry, first implemented in 1972 by psychiatrist Kenneth Colby.”
After that, chatbots remained under the radar of mainstream technology news for some time. But work on new ones didn’t stop, such as the one developed in 1994 by the founder of Lycos, Inc. (first known as “Julia”) to compete for the Loebner Prize.
Another example was an intelligent, chatbot-like customer service solution from 1999. It was the result of a partnership between FaceTime (then the leading provider of P2P interaction services) and Big Science Company (creators of the first application server to generate AI-empowered graphical assistants – Klones).
However, the technology that was then widely available was too immature for widespread adoption of chatbots. The markets weren't ready for them.
Recent history of chatbots – explosion in popularity
Currently artificial intelligence has developed to a point where programs can learn and effectively mimic human conversations. Accelerating technological progress has placed internet-enabled machines in every institution, company, home, and eventually pocket (who doesn’t have a smartphone nowadays?).
In this environment, chatbots have become increasingly popular as useful tools for companies and institutions. One of the best known examples of chatbots in recent history is Siri – the AI assistant that is part of Apple's standard software for its products. Siri took chatbots mainstream in 2011.
Since then brands in every industry have started to use them, eventually sparking a new trend – conversational UX. This refers to a User Experience in which your interaction with a company or service is automated based on your prior behavior (like working together with someone who is getting to know you).
If you're a programmer, you can take advantage of software like Alexa, which enables the use of voice to control devices.
On the other hand, if you are a consumer, you can already interact with chatbots on popular messaging platforms such as Facebook Messenger, iMessage, Skype, Snapchat, etc. According to Business Insider, about 60% of US millennials and Gen X adults have already done so!
An extreme example of this is Xiaoice, a chatbot that imitates a 17-year-old girl. It was released for a public test in 2015, on the Chinese service WeChat. Over a million people began talking to her within the first three days. Today she has had more than 10 billion conversations with WeChat users, she has 40 million followers, and over 10 million people have told her, “I love you.”
Healthcare and chatbots
The use of chatbots has spread from consumer customer service to matters of life and death. Chatbots are entering the healthcare industry and can help solve many of its problems.
Health and fitness chatbots have begun to attract a market. Last year Facebook has started allowing companies to create Messenger chatbots to communicate with users. A great example is HealthTap – the first company to release a health bot on the Messenger app. It allows users to ask medical questions and receive answers from doctors.
Currently, there are dozens of health and fitness chatbots available online. Fitness bots dominate this category, but there are plenty of medical bots worth attention too. Chatbots are far from widespread adoption in healthcare, but they are on the rise. Here are 14 notable examples of niche market chatbots:
Behavioral change: GoCarrot
Diabetes support: Brook.ai
Health assistant: Betterise
Medical second opinion: iCliniq
Medication adherence: Sense.ly
Nutrition coaching: Forksy
Oncology: Mendel Health
Patient engagement: LifeLink
Weight loss coaching: Lark
Tools for navigating healthcare: HealthJoy
Chatbots have been around since 1966, but their popularity didn't grow much until Siri appeared in 2011 and then FB Messenger bots this year. The market is constantly growing, with many startups that recognize the potential for using chatbots in health care to support patients and providers.
“Soon, patients will be able to enter their current symptoms through a portal, with the help of an intelligent agent, and get an accurate diagnosis or prescription without involving a human doctor.”
In the US, digital healthcare has a growing camp of supporters. One of these is Bradley Merrill Thompson, head of the CDS Coalition. Bradley argues for loosening of government regulations regarding automated medicine in the US and writes that “just as cars are beginning to drive themselves, healthcare decision-making is facing its own automation ramp up. Soon, patients will be able to enter their current symptoms through a portal, with the help of an intelligent agent, and get an accurate diagnosis or prescription without involving a human doctor.”
Technology has gotten smart enough that it can help resolve key problems in healthcare, such as a shortage of doctors – not only in the US, but on a global scale. It’s just a matter of implementing the technology.