Microsoft presented in 2014 its chatbot named Xiaolce, used on one of the popular Chinese microblog platforms, Weibo. The chatting application was programmed in a way that it chats with users like a friend. Shortly, for those who regularly had conversations and shared their personal issues with Xiaolce (having an artificial intelligence), it wrote “I love you” and “miss you” to the algorithm. With the help of machine learning, the developed chatbot was able to detect the feelings and current mood of the user just by reading their messages, and – if it felt like an intervention was necessary – it brightened their day successfully. Because Xiaolce – unlike real life users – was always patient and kind, remembered previous conversations and recalled certain topics.
The chatbots spread in large quantities in the beginning of the 2010s as a specific form of business intelligence. The interactive technology primarily based on artificial intelligence found a way to break into the world of online chat programmes, and nowadays users can encounter them on several platforms. But chatbots are not only parts of the virtual assistant. Several organisations and governments use it for websites, applications and direct messaging platforms in order to promote their products, ideas, services or topics considered important by them.
Chatbots found on different websites and applications can change the way how users communicate with governmental bodies, enterprises and institutions. Based on the estimations of Juniper Research, users sent 28.2 trillion messages in 2017, so we’re talking about the segment of communication which is very much in the centre of software and chatbot developers’ attention. Most IT companies want to join this revolution, therefore it’s not surprising that they invest huge amounts of money in the technology. Projects like Babylon Health interested in health care, Ozlo focused on gastro business and GoBtler interested in several segments of e-commerce invest 8 to 24 million dollars in x.ai, in other words, your future personal assistant.
But where does the engine of chatbots, artificial intelligence stand now? For instance, the chatbot named Eugene Goostman made everyone believe that it was human in 2014, because it managed to successfully personate a 13-year-old Ukrainian boy with limited English vocabulary. 10 out of the 30 judges of the Royal Society of London believed that they were actually talking with a living person. And the machines are more and more similar to humans. With the development of the technology (which basically means the progress of artificial intelligence, machine learning and natural language understanding), their use will be much more regular and intuitive. The goal of the algorithms is undoubtedly to take over the role of a real human operator.
The types of chatbots vary between simple and complex – the latter ones exploit a very wide range of artificial intelligence. On the other hand, simple chatbots handle the basic messages and requests of the users. For instance, they offer the information found in the menu of a certain website. The more complex chatbots can make the communication of governments and enterprises with citizens, clients much more effective. For instance, the chatbots used in the hospitals of the USA provide the opportunity of booking an appointment to patients. During the process, they just have to give the name of the doctor and the desired date – all of this is done in the form of chatting.
Chatbots also have a hybrid version, which – in case the algorithm cannot solve the task – involves a human operator in the conversation. The English Amelia works for the system of the e-government by automatizing customer service. The chatbot helps the dwellers of London get construction permission. If Amelia cannot answer a question reliably, she reports to her human operator colleague and then pays attention to the interaction so that with the help of the mechanical learning she’ll be able to answer similar questions the next time. There are significant developments going on in Hungary as well. Thanks tochatbots, users can ask for deliveries. With the help of sufficiently parametric questions, the chatbots takes into consideration the differences between districts before making an offer. For instance, it offers different meals in the business quarter than in the suburbs.
Another example of governmental chatbots is the development of MIT. The famous university developed a map system, which shows the regions affected by floods in real time. The system was first used to map the destruction caused by Hurricane Irma. The users of Twitter, Facebook and other social media platforms can chat with the chatbot of Risk Map in order to report any problems perceived. Then, the chatbot sends a link, which asks for a few simple data, such as location, deepness of water etc. and photos can be attached as well. The project attracts attention to the fact that information coming from inhabitants plays a crucial role in emergencies, and chatbots provide effective support in this process.
Next up, let’s learn about the chatbots of the e-government system. For instance, Visabot provides information regarding the immigration rules of the USA to help understand corresponding laws, while also taking into consideration the personal background and biography of solicitants. It also helps in filling out the forms that many times seem overly complicated.
Last but not least, I’d like to mention a noble initiation focused on Africa. To attract attention to the lack of drinking water in Ethiopia, “Charity: Water” developed a chatbot called Yeshi. Yeshi personates a young Ethiopian girl, who has to walk 2.5 hours every day to get clean water. The chatbot gifted with deep emotional experience plays upon geographic, online sharing and storytelling techniques. With her help, users living all over the world can learn about the struggles of people living in Ethiopia as if they were listening to a real human being.