Criteria for ‘Intelligence’
As I grew older and older, the meaning of ‘intelligence’ has been changed a lot, constantly. When I was a middle school or high school student, I used to call one my friends ‘intelligent student’ whom I wanted to be and in this context, intelligence meant an ability of person to merely memorize text book knowledge in a short period of time and get high score at an exam.
Examining the history of definition ‘intelligent’, we can see the origin of word started from 1508, indicating “A person who receives or takes in information; a listener”. It has been evolved slightly into “An intelligent person; spec. a member of the intelligentsia”, and “Having a high degree or good measure of understanding; quick to understand; knowing, sagacious”, and so on. The word, ‘intelligent’, has been used to cover a number of subjects within a number of contexts. Because of its ambiguousness and its broad-context-usage, researchers, historically, wished to concretize its concept into human scale: they have been made a huge endeavor to measure and reveal intelligence with quantified numerics, an IQ test was created. Surely, ‘intelligence’ was used for human-being for a long long time since entire decision makings were conducted by human beings and consequences of them were our responsibility.
It was until 1846 that ‘intelligent’ described only alive creatures, especially people, and it was that time when ‘intelligent’ first defined the attributes of device or machine. In 1846, in the American Penny Magazine, and Family Newspaper, Fitzgerald’s Braiding Machine was illustrated as an intelligent machine — “There has seldom been invented a machine, in which a greater effort of inventive genius was evinced..than in this the apparently intelligent machine”.
How come people started to call device or machine as intelligent things? An ability of intelligence includes receiving information, being rational, understanding, knowing, sagacious, communicative and etc. It’s common to say a person is rational, sagacious, communicative or knows a lot, understand well. But do machinery objects have all of those abilities? Are they able to heartfully understand something and make sagacious decisions? Of course not, they don’t have hearts to deeply empathize human beings. When we notice that Google Maps is tracking our movements or Facebook is collecting our activity histories thus recommending us things/people you might also like/know, we regard their ability as a mere ‘logical data processing’ rather than ‘understanding’ our desires. In my view, when people say “Oh, this device/system is intelligent”, it often means the logic of algorithm is well structured.
Therefore, there should be different criteria for intelligent things. I divided into stages of intelligent things: (1) receive huge amount of data (storage), (2) remember data promptly (recall), (3) apply pertinent data in certain context (operation), and (4) make successful/satisfactory decision based on understanding data (determination).
Apparently, we’ve seen exponential growth and evolution of artificial intelligence, started to learn the pattern of data flow and make decisions. AlphaGo has been a huge issue among us. It showed us a possibility of machine’s ability to learn things and judgements. Yet, we are still at an early stage of human-machine symbiosis. Still, we should take huge risks when we impart our responsibilities or jurisdiction to technology because of lack of ethical criteria, lack of understanding of various different contexts, and latent technical glitch. As designers, we should prevent speeding up imparting our jurisdictions to technology. Our role is to envision long term impacts of technology, keep iterating new technology or system before distributing it to globe dwellers and look back what intelligent means and what should it mean for our ecosystem and our descendants of descendants.
