On Human-Computer Interaction (HCI)

Interactive (Adj.) - capable of acting on or influencing each other

Interact (Verb) - act together or towards others or with others

Human-computer interaction (HCI) - the study of interaction between people (users) and computers. It is an interdisciplinary subject, relating computer science with many other fields of study and research. Interaction between users and computers occurs at the user interface (or simply interface), which includes both hardware (peripherals and other hardware) and software (for example determining which, and how, information is presented to the user on a screen).

Does a computer User interact with computer software? Is it an interaction??

Rodney Brooks from MIT claims that interaction between 1) a program, and 2) the world taking place during the computation, plays a key role that cannot be replaced by any set of inputs determined prior to the computation. In the case of artificial intelligence interaction can be viewed as a prerequisite for intelligent system behavior.

The concept of interaction entails some type of behavioral novelty. Intelligent behavior therefore respectively encodes and reflects the qualitative and quantitative aspects of the environmental states “taking place” during an interaction.

An environmental state is a function of a previous environmental state and the occurrence of one or more events.

State1 —> event —> State2

State1 —> event -> event —> State2

Note also that distinct causes have distinct effects, and as such there are many potential initial states that, upon the occurrence of a distinct event, become State2.

StateA —> event1 —> State2

StateB —> event2 —> State2

StateC —> event3 —> State2

Interestingly, using a thermodynamics-based model, upon reaching State2 one is unable to determine from which previous state one arrived.

In most all cases, a person (usually a domain expert) identifies those environmental states or changes that warrant or necessitate a system response. A computer programmer (who also may serve as a domain expert wrt software creation, testing, debugging, etc.) reflects on the pertinent environmental states, and distinguishing and symbolizing each for use in a computer model.

Computer software developed in such a manner lacks the type of intelligence described by Brooks. Why is that the case? The person who considers which environmental states warrant or necessitate a system response is in fact applying a qualitative assessment in determining such. And while such an approach explicitly accounts for the quantitative aspects of an environmental state, any qualitative considerations remain implicit, the person consciously focusing on distinguishing the environmental states quantitatively.

The issue resides in the way in which environmental states are symbolized. Those states that warrant or necessitate a system response, afford a qualitative sense to the person so making the determination. Only after such an affordance does the person ascribe a symbol reflective of the quantitative (objective) aspects of an environmental state. The symbol selected fails to encode for the qualitative considerations that led to the determination that the environmental state warrants a system response.

This qualitative regime must be modeled and incorporated into computer systems to connect the abstract quantitative symbols to the situation being represented. The qualitative aspects afforded by such a situation are what connect people to the environment in which they live. So too must it be the case with any sensitively adaptive intelligence.

Applying these concepts to web applications leads to the concept of Web2.U.

If you’re interested in this qualitative model please email me.

And as always, your comments are invited.

Much Thanks.