Virtual assistant It quickly became clear

[ritu789]

It took half a century to create a dialogue system capable of maintaining a conversation with a person on various topics. In the summer of 2014, the chatbot Eugene Goostman, developed by a team of programmers from Russia, passed the Turing test at the University of Reading with the support of the Royal Society of London. During a five-minute dialogue, the computer program convinced 33% of the expert interlocutors for the first time that they were talking to a 13-year-old boy from Odessa, Evgeny Goostman. Note that the percentage of specialists misled by the chatbot was even higher than the test conditions required, formulated back in 1950 by the famous English mathematician.



University of Reading Photo courtesy of University germany email list of Reading Professor Jane Setter Talkative Intelligence The brilliant result of this computer program demonstrated the readiness of the technology for use in everyday life. Other factors that accelerated the transfer of experimental developments into practice were the more active use of new digital channels by companies - websites, social networks, mobile applications for communication with potential customers and sales, as well as the growth of smartphone penetration and the craze for instant messengers. Telecom operators, retail chains, and banks quickly became convinced of the effectiveness of chatbots. And this is not surprising, according to various estimates, automated dialogue systems relieve 40% to 70% of the workload from employees consulting clients in "digital channels".



The system operates 24/7, it does not need a lunch break or an annual vacation. Large companies attract external developers or resort to the services of internal departments to create such computer programs for customer service. As a result, they receive highly productive automated dialogue systems, often capable of self-learning. that a virtual interlocutor - a salesperson, consultant, support specialist - is in demand not only by large companies with million-strong customer bases, but also by smaller businesses: online stores, ticket booking services, taxi orders, etc. Companies developing platforms for constructing chatbots appeared on the market. And popular messengers WeChat, Skype, Twitter, Facebook Messenger, Telegram, in turn, opened their APIs for connecting such computer programs.