Pages

Saturday, March 26, 2011

Artificial Intelligence and Humanity I

This post was based and inspired on the writings of Brian Christian


Turing test

The objective of the Turing test, created by Alan Turing in the year 1950, consists in evaluating during small conversations, if computers could imitate humans. The test incited the creation of Chatterbots such as ELIZA, launched on 1966. Later on the Loebner prize was created, offering a monetary award for the first chatterbot that could pass the test, this competition still takes place once a year.

The classic Turing test consists in one room with judges that make questions for 5 minutes to computers or to humans that are in a second room. After conversations take place judges reveal which chats they think they had with chatterbots and which with humans. The chatterbot that is considered the best is the one that fools the higher percentage of judges. Turing believed that the test would have been passed by the year 2000; far from his prediction a new record that was able to deceive only 30% of the judges was recently achieved.

Controversy

The principal question Turing wanted to solve with this test end up generating lots of controversy because it derivated mainly in three others "Can machines think?", "Are machines intelligent?", "Do machines have a conscience?" Lots of arguments where stated in favor and against them (including the ones that follow), but no unanimous conclusion was reached:

  • How we could tell if a machine was intelligent. After all, mankind has tried to define intelligence for ages and had made little progress except to decide that whatever it is, we've got it.
  • How do you know if the machine actually understands what it is doing, seeing, or saying? A particularly strange side effect of being a conscious being is that you can never truly know that someone other than you is conscious.
  • Not until a machine can write a sonnet or compose a concerto because of thoughts and emotions felt, and not by the chance fall of symbols, could we agree that machine equals brain-that is not only to write it but know that it had written. Neither until they can feel an emotion (Jefferson).
  • Computers can’t originate anything we explicitly tell them to do (Lovelace).
My preferred theory concludes that machines can't think; I argue that thinking cannot be separated from feelings (Damasio) and that it also includes an intuitive component.
Questions and thoughts
  • Do things have an essence, and if so is it possible to describe it? (Like Plato did in the Allegory of the Cave)?
  • What makes something to be something? We guide ourselves mostly by external appearances and actions. But indeed I think the evaluation would be more accurate if we compare that something internal structure and organization.
  • Can we extend the meaning of consciousness? By adding an exception to inanimate objects that does not include the act of thinking, emotions, intuition or any other aspect of the brain, in this case being conscious would mean only to have knowledge of their own existence.
  • Our brains consider that a certain entity is or is not something often by doing lots of assumptions. Instead we could assign a degree of credibility of what we think the entity is.