Guardian is reporting that six computer programs will answer questions by human volunteers at the University of Reading in a bid to become the first recognized 'thinking' machine, in essence pass the Turing test. To me it appears that it would probably prove nothing, at least for the time being.
For instance, if the “AI” computer programs are coded to follow a particular grammar and semantics, it can, never be taken as a machine that “thinks”. Over time, humans have evolved their grammar and language, can machines do the same? For the time being, I do not think so. A big example of this is the way SMS text messages are frequently sent [r u tr?, ts is gr8!, i wil cal u l8r!] … can a computer program that is only programmed (or “learn”) how to interpret English, would ever be able to “think” and understand this SMSenglish? I seriously doubt!
No comments:
Post a Comment