Being and being human. Solving the CAPTCHA.

Is it more human to correctly identify traffic lights in a series of photos, or to grow frustratedwhen we fail?

A CAPTCHA is a test designed to check whether the user accessing a website or filling out an online form is a human being or a bot, an automaton, a machine. That is done by presenting a test that only (and here the doubts and uncertainties begin) a human being can solve, such as precisely selecting images with a traffic light in it or performing, according to instructions, some mathematical-semantic operations. The CAPTCHA is a security check preventing unwanted access to resources, but it’s also a tracer tool that unsuccessfully attempts to draw that red (or rather grey) line that separates the machines from the human beings.

If I were asked whether, considering the approximate results of the latest generation AIs, it is premature to speak of blurred lines, I could only answer with a decisive no. The discussion must be exhausted before this line becomes blurred not only conceptually, but also practically. The topic must be articulated and stretched out in its most hidden folds immediately, while this rivalry in the solution of CAPTCHAs is just a meme and nothing to worry about concretely.

The acronym CAPTCHA itself is very exhaustive. It means: Completely Automated Public Turing test to tell Computers and Humans Apart. The tests used today are becoming easier and easier to bypass using neural networks for machine learning. Nowadays, on many occasions, a computer vision system is even more effective than a human in recognizing objects, text, numbers and symbols from image cutouts. For this very reason, next-generation CAPTCHAs no longer rely solely on this type of testing but on the source of the access request, its georeferencing, and its browsing history. That solves (partially) the practical problem of CAPTCHAs but still leaves the metaphysical one untouched.

Defining human essence has been debated for millennia. For Socrates, being human meant being endowed with reason, thus having the ability to think in search of truth through a critical examination of one’s ideas and beliefs. This process led, according to Socrates, to the development of a moral conscience, with virtue being the end and purpose of this search.

For Plato, the human being was definable as such for its refraction on a spiritual and metaphysical plane. It was certain that the purpose of being human was condensed in the presence of an immortal soul, separate from the physical body, which had as its ends the well, the beautiful, the good and, of course, the truth (not the bad). This divine breath present in every man represented the very definition of being human and was a means by which to differentiate oneself from the rest of things by attaining perfection.

For Aristotle, too, the understanding of truth was the core to the definition of being human. He thought human nature was defined by its reasoning skills and understanding, which distinguished it from other animals (and certainly from things). He believed that the essence of humanity consisted in the ability to know the truth and that the supreme goal of men was happiness, understood as a virtuous and fully realised life.

The undelying problem with these explanations is their gap in the direct definition of “reasoning.” Humans learn to reason — within cultural and instinctive patterns — but so too can animals, and increasingly, machines. Is their reasoning genuine, or merely a simulation?

This question was posed as early as 1950, in Alan Turing’s seminal paper Computing Machinery and Intelligence. Decades before computers reached their present capabilities, Turing anticipated our current predicament. He warned against the petty triumph of catching a machine in error, reminding us that humans err frequently, and that superiority is always relative — some men may outthink some machines, yet other machines may outthink them in turn.

If reasoning is insufficient as a dividing line, perhaps Descartes’ dictum still stands: cogito ergo sum. Extending Turing’s logic to its limit: “The only way to be sure a machine thinks is to be the machine, and to feel oneself thinking.”

On the triumph of the “correct answer” as proof of existence, humanity, and superior skill, Tuning wrote:

“Whenever one of these machines is asked the appropriate critical question, and gives a definite answer, we know that this answer must be wrong, and this gives us a certain feeling of superiority. Is this feeling illusory? It is no doubt quite genuine, but I do not think too much importance should be attached to it. We too often give wrong answers to questions ourselves to be justified in being very pleased at such evidence of fallibility on the part of the machines. Further, our superiority can only be felt on such an occasion in relation to the one machine over which we have scored our petty triumph. There would be no question of triumphing simultaneously over all machines. In short, then, there might be men cleverer than any given machine, but then again there might be other machines cleverer again, and so on.”

So, perhaps, for the discrimination between the definitions of machine and human being, we still need to use Descartes’ truth: cogito ergo sum. Therefore, taking the very concept of self-consciousness of a machine to the extreme, we continue Tuning’s argument:

“According to the most extreme form of this view the only way by which one could be sure that a machine thinks is to be the machine and to feel oneself thinking.”

It is not reasoning that makes us human beings: it is reasoning as human beings that defines us as such. Mechanistically speaking, human beings reason as human beings through calculations, signals and relationships of a biological nature. We have no way of ruling out the possibility of arriving at a simulation complexity such that mechanisms of human origin can reason as human beings. Self-awareness is not the certainty that we are not machines (am I not a machine? or a biologically encoded software? ) but the possibility of doubting that we are not.