5 Whose syntax and structure lend themselves to interactive use — синтаксис и

структура которых подходит под диалоговый режим ^в feedback —обратная связь

⁷ at the point of generation - в момент порождения ^B sprite—спрайт (растро/we графическое изображение небольшого размера

(напр.. 32 на 32 точки), которое может перемещаться по экрану независимо o-i

основного изображения) ^н schedule — график

81

by the year of 2000. It pointed to the fact that by the end of the century tbe whole of the Earth's population shoui'd be engaged

in programming.

Even a layman' can say that this cannot be true, the schedule has an error somewhere/Where is that error?

The schedule is based on quantitative data. Indeed, the time is not far away when computers are going to be part of our home, very much as it is the case with colour TV sets today. But in the near future computing science is sure to develop qualitatively. Consider the fact that when turning the TV controls² we never think about Maxwell's equations or single-shot multivibrators³ which form the foundations of present-day television technique. Similarly, we may not be aware of the prospects of computer development, namely that the latter are bound to grow more "humane" and "intellectual". It is not engineers alone that are concerned today with⁴ the problem of creating an artificial intellect.. This problem focuses the interests of researchers representing all branches of human knowledge without exception. Consequently, it is no wonder that³ this is an area which is expected to make qualitatively fresh gains ^б Creation of artificial intelligence is under way⁷ in the Soviet Union.

A reservation⁸ should be made at this point. The corresponding terminology is to be used correctly. You must have repeatedly come across expressions like "a case diagnosed by the computer". It is not the computer but the doctor who diagnoses the case, the computer just supports the expert's memory.

The World Health Protection Organization recently circulated-copies of case histories including numerous tests, cardiograms and other objective data concerning a number of patients from Scandinavia. Every copy carried a request to the leading experts to devise a correspondence analysis⁹ of the case. It is noteworthy that on the basis of the same data the experts would make not

' layman — непосвященный, непрофессионал

² when turning the TV controls — включая телевизор

³ about Maxwell's equations or single-shot multivibrators - об уравнениях Максвелла или лампах .бегущей волны

⁴ It Is. not engineers alone that are concerned... with — He одни только инженеры

заинтересованы в ³ it is no wonder that — нет ничего удивительного в том, что

⁶ to make qualitatively fresh gains — выйти на качественно новые рубежи

⁷ creation of artificial intelligence is under way—искусственный интеллект на­ходится в процессе создания ^в reservation — оговорка " correspondence analysis — заочный анализ

82

only different but often opposite inferences.' It came out that even a great number of formalized objective data. exact as they mav be² are not sufficient to diagnose a case with precision.

The point is that medicaUiagnostics ought to involve not only banks of data but also banl<s of knowledge Data and knowledge are not the same thing. Data are primary, passive information input into the computer. Knowledge is active, its contradictory nature makes man strive for overcoming it. Therefore the simulated model of the user's mind ought to be introduced into the computer's memory.

It is noteworthy that this possibility was provided for³ at the dawn of computerization by the Soviet-made computer “БУ CM-6”. A team of Soviet researchers headed by Academician S. Lebedev put into the architecture of the computer a number of fruitful ideas which remain topical- to the present day. Lebedev's team designed a system called “ДЖИН” serving as an interface to operator. In the process of inquiry “ДЖИН” could obtain data to build a "psychological portrait" of the user. The result was that “ДЖИН” would come to "like" or "dislike" the user.⁴ “ДЖИН” even "took offence". A questioning of the operators and program­mers showed that this activity on the part of the computer was not to their liking.⁵

Scientists are still puzzled over the exact mathematical simu­lation of the process of man's taking a decision. Cybernetics today includes the abbreviation PTD made up from the Person Taking a Decision. The PTD mental processes belong to the category which is hard to be formalized. Many things here depend on experience. As a rule, the solution ties in an expertise, i. e. Joint effort of a number of experts aimed at solving a definite problem. Such kinds of joint activities of several PTDs have long been used in medicine (conference of specialist doctors'}⁶ or sports (board of referees and judges).-

The technique of expert evaluation is widely used in determin­ing the strategic trends of research and engineering progress, in elaborating economic and social programs for years to come. A novel technique of organizing expertises was devised as early

' would make not only different but often opposite inferences — приходили не только к различным, но и к прямо противоположным выводам

² exact as they may be — какими бы точными они ни были

³ this possibility was provided for — зта возможность была предусмотрена

would come to "like" or "dislike" the user — стал “любить” или “не любить” конкретного пользователя

this activity on the part of the computer was not to their liking — подобная ак­тивность ЭВМ им не нравится conference of specialist doctors -- консилиум for years to come --- в будущем

83

as in the sixties' by Academician V. Glushkov. This technique has also found an application for solving the problem of planning allocations for² fundamental investigation.

"Our days see a regular boom in computer-based expertise systems," said professor D. Pospelov.

There are about a hundred expertise systems of all kinds func­tioning in the world. They are used in medicine, experimental chemistry, pharmacology, geology and archeology. Expertise sys­tems are being introduced in automated designs and research, in economics and history—in all cases where the specialists have to grapple with³ great masses of data which are not subject to any formalization.⁴

As distinct from data,⁵ knowledge cannot be just pumped over onto the computer magnetic disks from reference books, treatises and other clever publications. They can be fed into the computer by genuine experts. This input leads to the computer forming a kind of crystal semantic lattice⁶ whose joints remain blank. In time the blank spaces are to be filled with so-called frames⁷ which correspond to certain conceptions. A system results which possesses qualitatively novel properties. It reacts to changed situations in quite a definite manner. In case some frame cells are substituted for other conceptions the evaluation of a situation and the corre­sponding behaviour of the system change essentially. The method of frames, for one thing,⁸ is employed in systems of automated translation. It appears that even a super-computer is powerless in translating, say, an English text into Russian with one-to-one correspondence accuracy⁹ This cannot be effected even provided the computer memory contains all the words and phrases included in the dictionaries by Ozhegov and Webster. The point is that every particular language has a set of frames of its own, i. e. those elusive descriptions of typical situations which are stumbling-blo'cks¹⁰ for a foreigner.

A certain researcher has input to the computer 500 various

' as early as in the sixties—еще в шестидесятые годы

² allocations for — ассигнования на

³ to grapple with — справляться с

⁴ are not subject to any formalization -— не поддаются никакой формализации

⁵ as distinct from data — в отличие от данных'

⁶ crystal... lattice — кристаллическая решетка

⁷ frame — фрейм (в искусственном интеллекте — единица представления знаний, описывающая понятие как объект. Фрейм состоит из ссылки на родовое по­нятие (суперпонятие) и описания свойств, отличающих данный объект от су­перпонятия)

⁸ for one thing -- к примеру

⁹ with one-to-one correspondence accuracy — с абсолютной точностью

¹⁰ stumbling-block—камень преткновения

84

texts, scientific, publicistic, fiction and even jurisprudence. Every category of texts was found to be marked for an individual frame structure' despite the fact that all of them are made up of similar "bricks", i. e. characters, figures and punctuation marks, Moreover. every category is noted for a rhythm of its own, i. e. wavy semantic process in which certain dominant frequences can be observed. Thus a new approach is discovered to automated translation from one natural (human) language into another. If formerly enormous computer resources had to be spent on searching certain grammati­cal units in each sentence, for example, a verb or an adjective, analysis of textual semantic rhythm brings out the semantic aspect of translation.

Taking into account³ the topical³ character of the problem of creating such-kinds of systems for a number of purposes, the Presi­dent of the USSR Academy of Sciences has set up a special council to investigate the issue of "artificial intelligence". The council actively cooperates with various departments of the Academy as well as the Ministries responsible for the output of computer facilities.⁴

Today new scientific and engineering problems cannot be solved with the help of obsolete technique. A powerful stimulus of further advancement is the constantly growing intellectual potential of our supercomputers and robotic complexes of the fifth generation.