Know not how and why, but have been reading quite a few interesting articles about Artificial Intelligence in the past two days. Not so much the technical or mechanical aspect but the rational and philosophical dimension of AI. My general impressions and views, as of now:
What Artificial Intelligence (AI) has actually achieved is 'human intelligence'. It has only managed to create what human beings can think or do, better. But isn't AI about rational thinking and independent behaviour? The machines and robots have been programmed or coded with the various possibilities from the vast amount of available data, and generated algorithms thereof, to choose a particular act. It does not 'think' beyond the given data or possibilities. The created AI has only inherited the collated and multiplied human intelligence, but with its limitations as well. (Now this supposes a distinction between human thinking/act and rational thinking/act.)
Secondly, the inherited intelligence is that of a select group of inventors, not of all mankind. Example is the response of Google assistant, American made and Alisa, a Russian make to the statement, “I am sad.” Google's reply: “I wish I had arms to give you a hug.” Alisa's reply: “No one said life is going to be fun.” The former reply is typically a Western capitalist response to an emotional state and the latter a hard hitting fact of life reply as per Russian culture.
While consensus on what exactly does 'cognition' or 'understanding' mean and how it is achieved is elusive, to expect or attribute the same to AI will be even more complicated. Other crucial issues AI is yet to address: subjective consciousness, creativity, meaning… (Or is it that AI researchers and developers are not interested in mechanizing these essential features of the humankind!?)
What Artificial Intelligence (AI) has actually achieved is 'human intelligence'. It has only managed to create what human beings can think or do, better. But isn't AI about rational thinking and independent behaviour? The machines and robots have been programmed or coded with the various possibilities from the vast amount of available data, and generated algorithms thereof, to choose a particular act. It does not 'think' beyond the given data or possibilities. The created AI has only inherited the collated and multiplied human intelligence, but with its limitations as well. (Now this supposes a distinction between human thinking/act and rational thinking/act.)
Secondly, the inherited intelligence is that of a select group of inventors, not of all mankind. Example is the response of Google assistant, American made and Alisa, a Russian make to the statement, “I am sad.” Google's reply: “I wish I had arms to give you a hug.” Alisa's reply: “No one said life is going to be fun.” The former reply is typically a Western capitalist response to an emotional state and the latter a hard hitting fact of life reply as per Russian culture.
While consensus on what exactly does 'cognition' or 'understanding' mean and how it is achieved is elusive, to expect or attribute the same to AI will be even more complicated. Other crucial issues AI is yet to address: subjective consciousness, creativity, meaning… (Or is it that AI researchers and developers are not interested in mechanizing these essential features of the humankind!?)
No comments:
Post a Comment