Henry Sheppard: “In the future, information will turn into the most popular means of interface”

    By the ICQ number 311826826 a special guy Serega jumps out - Balagur, who, instantly joining the conversation, pits you a dozen or two jokes, shares personal and almost sells an elephant. Serega loves aquarium fish, saves for an apartment, and all requests to borrow 100 rubles are categorically answered "I am a humanist." With such a blue-eyed aquarius, you can’t go into intelligence, but don’t go anywhere at all, because Serega is a bot of the family of virtual interlocutors, the info that lives only on ICQ and on the website of the Nanosemantics company that gathered his “fathers”.

    Intelligent bot became known from Igor Ashmanov, CEO of Ashmanov & Partners, who said that Balagur is a beta version of the infa, sharpened for entertainment of the interlocutor. “The thing is modern, quite a lot of labor has been invested in technology,” Igor says about the bot. - People talk with him, make fun of - who scolds, who likes, but there are sessions and 500 replicas, that is, 8-10 hours . In general, addictive. "

    "Habrahabr" talked about robots with Henry Sheppard, the head of development of information in the "Nanosemantics".

    Please tell us about the technology for developing infa. How does it work?

    Immediately, pure mathematics and logic have not yet been revealed. We use fairly advanced mathematical methods that have not yet “shone” in similar projects, so we try not to spread too much about them.

    With linguistics, it’s simple: the scheme of work is known to everyone who has encountered the development of simple virtual interlocutors. A “knowledge base” is compiled containing a huge number of templates (there are a lot of problems that create the excessive complexity of the Russian language) that allow you to correctly choose declination, conjugation, etc.

    The mathematical part allows you to track semantic words and "hold the focus of attention" infa, so that he can conduct an adequate conversation with the user. Our inf can be not only a passive "reflector" - he sets the conversation topics himself or changes them if necessary. This is what distinguishes it from other bots - the “human” ability not only to respond “to the cashier”, but to constantly change the subject without breaking away from the general outline of the conversation.

    How is the "knowledge base" developed?

    If we drop the math, the “knowledge base” is a very crude term. It is conditionally divided into two parts: the template mechanism (the correct preparation of phrases) and the knowledge itself. Here the most interesting lies.

    The fact is that the creation of an “almost real brain” is impossible for common reason, but for practical reasons it is pointless. It is enough for INF to put some knowledge into it: for example, a simple joker is full of jokes to failure - this is convenient for demonstration. The set of knowledge can be anything - it can be the knowledge of a specialist administrator of the network, and the data directory of cell phones. In the first case, we get the perfect admin who can’t avoid the stupid questions of an accountant, and every time clearly explains everything. In the second, there is a sales assistant in a store who is never mistaken and can quickly find a suitable device even for a buyer who is “neither be nor me” in cell phones.

    That is, Balagur is the “first sign” of your company, followed by others with concrete practical applications?
    Practical - are already being developed. But customers do not really advertise the "robotic nature" of their communications. Indeed, as a support service, inf is a small and very economical miracle. It is easy to guess that the customer doesn’t really want to write on his website, they say, consult with an artificial interlocutor. Inf sometimes behaves even more humanly than a tired admin, so it’s logical not to advertise the artificiality of his “brains”.

    And Balagur is just a demonstration of technology.

    In the future, especially with the advent of good speech recognition systems, information will turn into the most popular interface tool. After all, keyboards, buttons, menus cannot be compared with the ability to give a command in a natural language.

    Henry, can a robot (in principle) be taught to think?

    In principle, no. The thought process is based on the ability to make choices in conditions of lack of data - this is unattainable in the absence of "analog processors". A modern machine, however, performs a clear algorithm, even if a program is being written to solve a problem with a lack of data. And the program itself is executed rigorously. One must add to this the fact that so far there is not enough information about the work of the brain, despite the fantastic achievements in neurosurgery.

    Personally, I, as the most severe materialist, are sure that it is possible to identify the mechanism of the brain. But this, let’s put it this way, is “a task that will eat up all the resources for its solution”. If you measure the temperature of the body by touching it with a thermometer - you need to make an adjustment to the fact that the thermometer also has a specific heat. Can he measure the temperature of this thermometer with another thermometer, and that one with another, etc.?

    Maybe making a robot too humanized does not make sense, because humanity is reproduced in a natural way .... Why strive to make full-fledged similarities, if the original is in abundance?

    Yes, but this is a matter of ethics. I personally have nothing against it, but a moral conflict will arise - whether to recognize them as personalities. But in a practical sense, this is not necessary. Robots should make life easier for humans, and not replace them.

    What is the reaction of people communicating with Joker, you observe?

    The techies, of course, are trying to “catch” him, which is easy - this is demo. But especially impressionable girls (and there are many, as it turned out!) Even try to invite him on dates!

    Regarding technology, you, as a developer, what are you proud of? Can you briefly determine its innovativeness?

    She's just one of a kind. And in a few years she will “bury” all these phones, computers, PDAs, etc. It all comes down to a small device with built-in info, which itself will go online (the Internet will turn into a protocol for exchanging and storing data) and find the necessary information for the user. In fact, it turns out a personal all-rounder.

    How many people freed out info Serge?

    Judging by the logs, maybe for a thousand, although we have not advertised it or promoted it anywhere. Now we’ll fix some little things, and then we’ll launch it as a separate project with the user’s ability to embed it on sites. So, a trifle, but nice.

    Also popular now: