[ CogSci Summaries home | UP | email ]

Grudin, J. (2009). AI and HCI: Two fields divided by a common focus. AI magazine, 30(4), 48-57.

  author = 	 {Jonathan Grudin},
  title = 	 {AI and HCI: Two fields divided by a common focus},
  journal = 	 {AI magazine},
  year = 	 {2009},
  volume = 	 {30},
  number = 	 {4},
  pages = 	 {48-57},
  month = 	 {Winter}

Author of the summary: Danielle Coleman, 2012, danielle13@hotmail.com

Cite this paper for:

Considering that both Artificial Intelligence (AI) and Human-Computer Interaction (HCI) explore the realm of computing and intelligent behaviour one could assume innumerable crossovers between these two fields – however this does not prove to be true. This article written by Jonathan Grudin attempts to outline the history of these two fields with the goal of explaining why, given the similarity between these fields, they are so divided. In summary the key differences between AI and HCI can be marked by the ambition, long term goals and expense of AI compared to the innovation and short-term time scale of HCI. These key differences lead to differentiation of methods and approaches by those in the field and ultimately AI and HCI began competing for time, economical and intellectual resources partly because HCI thrived on resources that were freed when interest in AI declined.

Establishing a Timeline

1950s: The potential of computation was realised in its role of code breaking in World War II. After the war the government funded the building of computers at specific universities. In the early 1940s and 1950s the Macy Foundation sponsored a series of conferences bringing together mathematicians, psychologists, and social scientists of which a couple of the topics discussed were cybernetics and neural network models, a begining of a breakthough into the research of both AI and HCI. The term ‘artificial intelligence’ first appeared in a workshop in 1956 written by American mathematician and logician John McCarthy.

Early 1960s: With the spotlight on AI the Managers of MIT’s Lincoln Labs looked for new users for their governments funded TX-0 and TX-2. That new user was found in Ivan Sutherland who constructed Sketchpad with introduced many interface concepts. This had a profound impact on the research of HCI and shaped what it is today.

Mid 1960s – Mid 1970s: One of the first downturns in AI was between the mid 1960s and mid 1970s with the launch of the Sputnik satellite in 1958 which refocused attention on scientific research. This downturn was short lived as AI started to take shape as a major research field starting with an influential essay written by J. C. R. Licklider titled “Man-Computer Symbiosis” which defined a major role for AI in exploiting computers.

In the 1970s Nicholas Negroponte of MIT argued that for machines to understand the context in which they operate they must be able to understand speech. For computers not to be able to do this they would be dangerous and unreliable, he stated. With this, copious amounts of funding became available to both AI and language-processing. Although 40 years later computers still cannot do what we would like them to do. During this time however research and man-power was devoted to improving screen layouts, command names, text editors, graphical interface and other HCI related advances.

However, during the mid 1970s overconfidence in rationality lead to HCI competing for funds and student interest against the exciting and important vision of AI.

Mid 1970s –Early 1980s: Eventually it was clear that AI had been ‘oversold’ which lead to the growth and development of HCI laboratories in such places as PARC, IBM, Digital, the U.K. Medical Research Council Applied Psychology Unit, Bell Labs, and the University of San Diego. The Human Factors Society’s Computer Systems Technical Group also formed during this time.

1980s: In the early 1980s a new foreign threat breathed life into the funding of AI. Other factors contributing to this bloom in AI such as when Symbolic and LMI marketed “LISP Machine” optimized for the western AI language of choice. At this time people started to echo Turing’s claim that machines would soon rival human intelligence.

1990s: This was the beginning to a long AI “winter” where funding was cut because progress in such area as autonomous land vehicles, battlefield management systems, and natural language processing seen no new developments even with the 2 billion dollars in funding that was provided. Although this way an AI “winter” this was a period of growth for HCI, many computer science departments added HCI to their core curriculum and hired HCI faculty.

Current Decade and Conclusion: Over the past years there have been notable achievements in the field of AI such as Deep Blue defeating world chess champion Gary Kasparov, remote controlled robots exploring both terrestrial and extraterrestrial space and autonomous land vehicles actually making it to the finish line at the DARPA Grand Challenge. Although these are amazing contributions to the field of AI they are overshadowed by the forecasts made by professionals who promised genius machines that could solve the world’s problems – from this people begin to lose hope in AI and turn their resources towards other fields, such as HCI.

This is the relationship between AI and HCI, where one field feeds off the other. When AI is up and blooming HCI looses funding, when AI is down and in its “winter” HCI flourishes – because of this these fields remain at arm’s length which makes crossovers seem like an impossible goal.

Summary author's notes:

Back to the Cognitive Science Summaries homepage
Cognitive Science Summaries Webmaster:
JimDavies (jim@jimdavies.org)