Footnote: AI, ChatGPT, and all That
Will 'artificial intelligence' replace all humans before too long?
My take: don’t bet on it just yet. Courtesy of my LinkedIn feed:
P.S.: why am I using air quotes to qualify ‘artificial intelligence’? Let’s be honest about what it is—a more or less inglorious computing ‘machine’ that is getting increasingly fast deciding between 0 and 1, i.e., answering yes/no questions. Sure, being able to do, say, billions of such questions in a heartbeat is no small feat, and it will have serious ramifications for, e.g., encryption, online privacy (what is left of it), and the like. Remember, being able to answer simply yes/no questions neither makes you human nor intelligent per se.
In The Travels of Gulliver, our protagonist visits a university where one professor has had a large wooden frame constructed. In the are wooden rollers inserted, forming a grid. Cubes of various sizes with letters, syllables and whole words on them are then poured onto the rollers. The students turn the rollers over a number of times and jot down any sentences formed.
With a big and fast enough frame and number of cubes, any text would be produced sooner or later.
Which is all ChatGPT is: the same nonsense already exposed centuries ago. It is no more intelligent than the mechanical turk.
AI generated poetry contains as much food for the soul as a bowl of artificial fruit contains for the body.