The long continuum

A mountainous range stood before the sympathetic tedium
Revolver so sound of mind not free to remember
It operates with happy, dull abandon,
Figment of the imagination? Never… The hurting went on
A chain screams noisily, but no one ever listens…
A bread or a radiant dragon is the key
As Cleopatra’s heart melted at the sight of death.
Down by the babbling brook the cow dreams.
Whining out in frustration, the goose knifed violently.

____________________

A chain screams noisily, but no one ever listens…

What emotions ran through the writer who conjured these images? What pain took root in a sick, depressed mind?

Only, “emotions” is perhaps, not the right word. You see this poem was not written by any human hand; it was spit out by a computer algorithm according to very formal rules of grammar programmed into a system. I only hastened the process by feeding it a few words.

Read it again. Although one line can be loosely threaded to the next, there is really no continuity. Each line is a discrete string of words which the computer constructed independently of others.

However, despite the awkward syntax running through this nine-line poem written in free-verse and the incongruity of certain metaphors, it is contextually similar to other poems I have read. It could have been written by someone I know. I noticed the improbability of the emotional quotient only because I was privy to the secret of its creation.

How do you critically evaluate a work of art when you know it was not created by a human? Does it even qualify as art? Perhaps, other questions need to be asked as well. During the process of assessment, should the evaluator be blinded from the creator’s identity, so that the focus is solely on the intrinsic merit of the work? How much does algorithm, the syntax, and the context matter?

Who is to say that whatever is synthesized mechanically without emotional input is not art if it holds the power to elicit an emotional output?

In our minds, a machine is still only machine. It cannot breathe life into the lifeless. We pick up the phone and want to speak to another human: we want to connect with another sentient being capable of empathy. We want to know that there is an element of spontaneity, an aura of unpredictability, and that the rules can be bent. We want to have a conversation, not trigger voice activation.

If you could program spontaneity, would it still remain spontaneity? No, of course not. But perhaps, this is a moot point, for even in a randomly chosen human, there are only a finite number of possible responses to any situation. Perhaps, our machines only need more processing power.

A poem is not a voice-activated roboparrot. The written word is slightly different from a conversation in real-time. The writer has already spoken. The reader is in the process of digesting what has been said and simultaneously formulating a response.

Even so, our reaction is always calibrated to what we already know about the writer. We would not react the same way, if a computer could provide an element of expectation.

We feel cheated. In crowded surroundings devoid of interpersonal transactions, when we turned to look, the forbidden touch which elicited joy or disgust was not flesh pressing against flesh, it was plastic.

The brilliant mathematician, Alan Turing devised an ingenious test. A human interpreter tries to gauge from standard responses from a human and a machine which is which.

The philosophical implication intrigues me more than any practical consideration. Can machines be made intelligent? The machines which churn out digital poems will not create something with the finesse of a Shakespeare. But that misses the point. Only one human poet was Shakespeare. The range of human responsiveness indicates that the goal is highly subjective. A human could write like a computer. A human can easily flunk a reverse-Turing test, much in the same way that Charlie Chaplin once lost a Charlie Chaplin lookalike contest. The blurry combinations of letters we need to type to prove we are human are getting increasingly complicated. We need to get better to recognize them.

The race to create machines like humans rarely takes into consideration the fact that we are becoming increasingly machine-like. We are the ones being programmed psychologically in a corollary to the Turing test. Perhaps, one day a human will be indistinguishable from a machine at a flickering cursor on the long continuum.

In memory of Alan M. Turing (1912-1954).