The First International Joint Conference on Artificial Intelligence is held in Washington, DC

A new venue, and a volume that records it

The First International Joint Conference on Artificial Intelligence (IJCAI) was held in Washington, DC, from 7 to 9 May 1969^[1]. Its proceedings were edited by Donald E. Walker and Lewis M. Norton^[1]. A proceedings volume is the collection of the papers accepted for a conference, printed as one book. Because it is assembled at a single moment, it works as a snapshot: it shows what a research field was working on at that date, and what it called its subfields.

Man-machine symbiosis comes first

The volume's first section is on man-machine symbiosis in problem solving, and it opens with Warren Teitelman's "Toward a Programming Laboratory," on pages 1 to 10^[1]. The word in the section title, symbiosis, describes a division of labour: a person and a program each take the part of a problem they are better at, instead of the machine only carrying out a computation the person has already fully specified.

Heuristics and theorem proving

The section on heuristic problem solving carries C. Cordell Green's "Application of Theorem Proving to Problem Solving," pages 219 to 240^[1]. A heuristic is a rule of thumb that narrows a search, so that a program does not have to try every possibility in turn; much of the work in this period was about making search feasible rather than exhaustive. The theorem proving section carries Carl Hewitt's "PLANNER: A Language for Proving Theorems in Robots," pages 295 to 302^[1]. A paper about a programming language, filed under theorem proving and addressed to robots, is one sign of how the volume is organised: logic, languages and machines appear together rather than as separate specialties.

Robots that move and see

The section on integrated artificial intelligence systems carries Nils J. Nilsson's "A Mobile Automaton: An Application of Artificial Intelligence Techniques," pages 509 to 520, and "The Stanford Hand-Eye Project," with Jerome A. Feldman and colleagues, pages 521 to 528^[1]. The titles point to machines that move, and to machines that link seeing with handling, so the section is about complete working systems rather than isolated methods. The label on the section is itself informative: an "integrated" system is one in which perception, reasoning and action have to function together.

Language, patterns, and modelling

The linguistic section carries Roger C. Schank and Lawrence G. Tesler's "A Conceptual Parser for Natural Language," pages 569 to 578^[1]. A parser is the part of a system that recovers the structure of a sentence from a string of words; calling this one conceptual indicates that it works with meaning-level representations rather than only with grammar. Other sections covered pattern recognition and signal processing, pictorial pattern recognition, self-organizing systems, physiological modelling and psychological modelling^[1]. That list shows how wide the borders of the field were drawn in this first volume: pictures and signals, self-organising systems, and models borrowed from physiology and psychology all had a place at the same meeting.

The series that followed

IJCAI records that its conferences were held biennially in odd-numbered years since 1969^[2]. The second IJCAI was held in London in 1971 and the third at Stanford in 1973, according to the official proceedings list^[2]. Washington therefore set a rhythm rather than hosting a one-off gathering, and the date 1969 became the starting point by which the later meetings are counted. For anyone reading the field's history, the first volume is the place where a set of names, problems and page numbers can be pinned to a particular three days in May.