What defines humans and computers? In need of a new Turing Test

Human body with a computer head standing within skyscrapers, digital art

When I was giving the title to this article (usually the most daunting part of writing), I struggled with which of the two, defines and demarcates, is a more fitting word to this article. It turned out that prior to a clear demarcation, we need to define what is so unique about humans that sets us apart from machines and any other kind of living animals. But later in the stage, I realised the definition of humans has never been more flimsy.

So what defines human?

Humans, or Homo Sapiens (Linnaeus, 1758), oftentimes are viewed as the very few that possess wisdom (or intelligence) and demonstrate the ability to do a wide range of simple and complex tasks. As suggested by the formal latin name of Homo Sapiens, sapience perhaps is one of the unique characteristics of our kind. As antique as it can get, Renaissance humanist philosopher Giovanni Pico della Mirandola in his book Oration in the Dignity of Man describes human as the most fortunate of all living things as we are “of the very intelligences which dwell beyond the confines of the world”. However, as we can see from real-world and lab animal observations, humans are not the only ones that have wisdom and are actively using them, some other intelligent animals such as elephant and dogs are shown to have the almighty creativity and wisdom, as Mirandola put it, to transcend our very physical realm.

Psychologists and biologists have also, reasonably, turned to the concept of sentience – being conscious, having mental states and ability to feel (psychological) pain. Perhaps one of the most defining book on the nature of human, which puts sentience as the crux of one, is Microcosmus: An essay concerning man and his relation to the world (1856) by the German biologist, medical doctor and logicist Hermann Lotze. On human nature, the author considered sentience as the most significant human characteristic and temperament- the capability to feel and the to be affected by feelings through the expression of one’s feelings in movements, words, and gesture. Well, I have seen my indoor European shorthair being scared of a larger-sized outdoor one multiple times and she sought after my comfort every time. Then how different are we from animals, just on the basis of sentience?

The answer might have been offered by the books uncovering the development and the so-proclaimed “truth” of human. The bestseller The Dawn of Everything by David Graeber and David Wengrow wittingly (and possibly implicitly) illustrates the uniqueness of humans does not lie within an individual, but the products of a collective of humans coming together – society and politics. Humans do not merely differ from other species because we are sapient and sentient, but also because we evolve both biologically and politically.

Humans were the only agency

It was always assumed that humans were the only ones who were able to respond the environment without, or by the least with little, instinct. Marx’s conception of agency echos much with Mirandola’s idea of free will. In other words, we are an agency because we are able to make free choices in a given environment (beware that in the Marxist theory of agency, humans are, again, social and collective being). Despite that we are the only agency, humans are always in the tug-of-war against the social systems and extra-biological structures of gender, social class, nationality and most probably the intersections of these.

While the Marxist approach to human as an agency holds some water, what we might have missed out is the existence of structures does not just impede the exercising of humans’ free will, but the structure per se is an agency. Such position is taken and amplified by STS scholars such as Bruno Latour and Peter-Paul Verbeek, as in their work concerning Actor Network Theory (ANT) and agency of technological artifacts. Look at the following sign:

Speed Limit (Adviessnelheid)

You will most likely slow down your car when you see this sign. This is a simple example laid out by Verbeek (2005) of how artifacts as crucial actants in our society shape our behaviours. The centre of their philosophy is that the asymmetry and de facto juxtaposition of humans and nonhumans must be annihilated to untangle the intricate web of our society.

Extend the bridges: The New Turing Test

But we have now come to phase where it becomes growingly challenging to even contrast humans and nonhumans, specifically Artificial Intelligence.

On the intelligence level, a myriad number of narrow AIs can already supersedes human-like intelligence, the Stockfish 13 chess robot for example. On the sentience level, Google’s LaMDA is thought to be the new breakthrough, though it was purely an extremely strong language model. If computers are shown to possess both human level sentience and sapience, what even are we talking about humans versus nonhumans.

The public might have perhaps laid too much emphasis on the inner processes and differences between humans and computers but have neglected what really is the core of challenge of the state-of-the-art technologies. This is revealed by the genius about Alan Turing. The Turing Test was designed to test whether computers (or artificial intelligence) can ever imitate humans in its 3-person guessing game of who-is-the-human and so far there has not been any success. The Test was, in contrast to public’s fantasy, not to investigate whether computers can think (sapience), but to focus on the output, whether it can pretend to be a human through various modalities.

Ultimately, from my own perspective, the importance does not lie within the processes of AI’s “sapience” or “sentience”. The key to a more comprehensive understanding of human and nonhumans, is whether sapience or sentience that are expressed by nonhumans can be disambiguated by ourselves. Potentially, a new Turing Test has to exist to examine humans’ perception of not only bots’ sapience, but also their sentience.