Today we are on the threshold of the revolutionary changes in the sphere of the communication between people, between people and machines, between machines. At the threshold of a new era: Internet-of-people (IoP).
I combine the Internet of people (iop) and the Internet of things (IoT) into a general notion of Convosphere (thanks to Chris Fine for offering this word). This industry, worth billions of dollars, has been rapidly growing since the dawn of social networks. Now having got a shot in the arm from the IoT, it is transforming into a new combination of the communication technologies, such as: chatbots and speech interfaces upgraded with a constantly improving artificial intelligence (AI) and communication platforms of a smartphone type.
Until today, we have not wondered who we are socializing with online: a human or a bot Now it’s time to think about it. Until now we have socialized with humans, but as chatbots advance we’ll have to communicate more and more with machines. Planning a rout from point A to point B we say “Ok Google” or «Listen siri». This area will expand further on.
Unfortunately, our communication on the Internet (email or social networks) becomes littered with spam. We subscribe to friends, acquaintances, news of online stores, and then waste time reading, browsing, and … unsubscribing. This is because after a while their news does not interest us anymore.
I get over 100 letters a day, 80% of which are newsletters, I am not interested in, or spam.
I do think that the present advertising of products or services is archaic. Personally, I buy only what I need, and I do not want to waste time decision making. I believe, a chatbot could do it for me.
Wanting not to look a dreamer, I still believe that we need personal robots-assistants such as a customized «Siri» or «Google», which would read for us emails, advertisements, and help us make decisions.
For example, I want to take a kid somewhere on a weekend, so I ask my robot-assistant to find suitable events and recommend the best. That would be very practical. “Ok Google” or “Siri” give you a list of events without analyzing if it might suit you or not.
My chatbot will know my tastes and preferences and will help with the right selection.
These are functions a personal robot-assistant might have.
Make smart Google search. You write a question, and the chatbot searches and gives you a relevant answer.
Monitor currency rates, stock market quotes and share market. E.g. “What is PTC index today?” “How much is an oil barrel?” or «What is the dollar rate?”
Follow the latest news on the sites you trust.
Know the weather in your area. E.g. What is the weather like in Moscow now?
Give quick access to entertainment sites. E.g. Where to go on the weekend? Current events guide.
Follow news on various topics: motorsports, basketball, safety, exhibitions, sport, games, movies, crime, culture, fashion, music, science, property, society, politics, law, events, religion, software, tennis, technology, football, hockey, business , ecology, economy, energy.
Make user-friendly search by Wikipedia. E.g. What is a boson? What is a quark? Who is Pushkin?
Help find the necessary info about authors, books, movies. E.g. Who wrote Hurry Potter? Who sang “Yesterday”?
And many more.
A chatbot could help you organize your activities. Your friends and colleagues could add reminders to your calendar.
Book an appointment. People can book an appointment with you via your chatbot. It will be automatically added to your Google calendar.
It is possible to expand the list of smart functions endlessly, to improve communication.
Monitor the info about actual discounts. It is important that you will get discounts when buying something (just ask your chatbot, “What discounts do I have in Big Lots?”).
Calculate your commute time. Going home from work, you inform the bot about it so that if your home people want to know when you are back at home, the chatbot could tell them the time.
I believe that in the nearest future chatbots will become our virtual extension.
Today, emails and posts in social networks consume much of our time. You will never know if you need the information contained in the messages you get until you read them. The chatbot will open an opportunity for your friends to add any information for you there, where you will be able to get acquainted with it when it is convenient for you.
Depending on the access settings you can share structured information, such as:
With the help of the chatbot it is also possible to perform certain operations without using other applications, such as: a phone or email. The chatbot, for instance, will be able to book a table for you at a restaurant or order a takeout, choosing one of your favorite places saved in the chatbot memory.
For sure, I’m not alone in the belief that the introduction of chatbots will save us from the applications dependency and change the way we interact with the Internet. Websites will lose some of their significance. The data we want to share, both for personal and public use, will be stored in the cloud. Our chatbots and their associated software agents will know how to collect and sort the information we are interested in, how to help us spend our time more productively. We spend billions of hours interacting with inefficient sites and obsolete communication systems such as email in an attempt to collect and sort out our information. Addition of the powerful data analytics and AI to the existing mixture will allow these new services to process huge amounts of information in order to most effectively meet our requests and save our time as much as possible.
A user-friendly interface is needed to create a chatbot.
Modern chatbot platforms are an excellent tool for developers, but they are completely unsuitable for end users, especially for private consumers who want to make their chatbot.
The main problem is that for this it is necessary not only to understand complex logics and settings, but also to know programming languages, such as Self, for example.
For chatbots to become popular, it is necessary to upgrade the platform for their development, making it very simple, practical and understandable. Although the process of creating a simple bot does not take more than 10 minutes, it requires special professional knowledge.
We believe that the Botlibre platform will make a very good basis for practical and understandable interface for bots development.
The memory and everything associated with it is the most important thing to furnish a chatbot with.
The easiest way to create a chatbot memory is to use the embedded neural network, which dynamically creates objects, linking questions and answers. But such a system cannot be used to memorize knowledge, because the user has few opportunities to fill the chatbot memory/knowledge with more structured data.
In other words, a more universal algorithm is required, according to which memorization would be a more accurate process without knowledge confusion.
Where * is a thought or a phrase to be remembered.
Where ** is the most relevant query to the memory cell.
To make the access to the knowledge base easier and more convenient, it is necessary to add to each element of the memory cell
analogies, synonyms, similar concepts, sets of keywords, incl. derivatives, as well as previous requests to fix (build) the search context;
in addition to keywords, necessary tags can be added so that «the memory would be more accurate».
The search is carried out using keywords and associations. The query may not have the right keywords or tags, but the phrase to be remembered may contain a full scaled number of the closest analogies. Thus, when memorizing a particular event/phrase, an object of the closest analogies is created, which allows you to remember the essence of the spoken/written info and get the access to this knowledge, even if the sought words are not there. For example, the memory contains the phrase: «Mom cleaned the window». So you can get the answer «Mom cleaned the window” to the question “What did a parent do?” because “a parent” is the closest analogy to the word “mom”.
Thus, knowledge, memory cells, i.e. separate elements to be memorized, must be stored in the memory with the help of word analogies. In addition to analogies, it is planned to add semantic analogies when a class of synonymous expressions is added to the thought, which is to be remembered.
What bots do we expect in 2017?
It should be mentioned that today we live better than the monarchs of medieval Europe. We have the internet, WiFi, hot and cold running water and, most importantly, the opportunity to order pizza via Telegram, but we always want more.
As any human being on the Earth, I’m glad that machines and computers make our life easier, safer and more comfortable, but I think that the real era of machines is yet to come.
The future, when we will communicate with robots and have robots as our friends and assistance is still in store for us.