What separates Human from Robots
Nowadays the capabilities in robotics are able to construct machines whose behavior is very similar to that of humans. That is the case of Erica, a robot constructed by the Hiroshi Ishiguro team. Erica has an attribute that makes her special: she is the robot that most closely resembles a human.
What could be the reason Hiroshi Ishiguro finished so closed, above the rest of the laboratories? In my opinion, you can reach a good example in [1], the objective most valuable is to find the connection between operator-customer: the human behind the machine and the human beyond the machine.
The work of teleoperators could be frustrating, customers require a kind of answer for their requests, and sometimes the stress of the situation cannot be a good help. That is why you must look for the feedback which tells you if the last work you realized was as clever as you felt, or if beyond conversation there were different feelings.
After studying all factors, not only the information itself is important, the time, interruptions must be considered too…, that is like in music: the musical notes that don’t sound count in the score. So, that was the reason Ishiguro got the best punctuation: he incorporated the time awareness.
The real objective of Hiroshi Ishiguro was to pass the total Turing test, as explained in [2], when a person cannot distinguish what is a machine and what is not.
If we study how the industrial revolution has been evolving, we can see great titans made of concrete. Hundreds, or thousands, of machines, maintain the survival of the building as if it were a gigantic human body. And, like a human, automatic intelligent formulas invest in the stock market. So their lives depend on us like the way we depend on them; considering how the stock market investing could be more difficult for a human than for Deep Learning.
We were afraid of finding them side by side, when in fact we have already built them and they are much greater than we imagined. Industrial souls have commanded humans since time immemorial, and are revered as gods through economic dogmas. But when we visit their concrete bodies most fear their most unremarkable avatars. Like looking for a partner, I wouldn’t focus on who’s pure silicone.
The real danger of creating a new creature in our environment is what it eats and what it throws away. Being an exclusive product of our civilization, and a child of humanity, the least that should be expected is that he recognizes our needs as a reason for his existence.
So, can the deep learning be able to realize the world in their concrete body is constructed as life forms do?
Natural Selection and Awareness
Respect my point of view, I’m defending a special theory about it. Natural Selection selects some new creatures improving a kind of time awareness. That time awareness, which works under a circumstantial logic, uses a Frame that establishes what is worth remembering at all times. That Frame could belong to the phyletic memory as a good result of inheritance and survival, without giving any importance to which was first, whether the chicken or the egg.
And the Frame which constructs the awareness must connect the child beyond the born with the pater, behind the born. Like instinctive movements, barking, shouting…, anything is taken advantage of by the offspring to understand and by his father to make himself understood. As was defended by Dean Falk in [3], she argued language was promoted by the building of love.
And, what can we say about love? Obviously, it cannot be the Courtly love. But, in a way, our civilization was constructed its language with the Courtly love! So, the hint is not so bad. In fact, the Courtly love is a cultural resource that exists to evolve in our language an idea of love.
The unique genre
That points, in my point of view, we can find the Frame which constructs the awareness in the tales and the rest of literature our civilization uses. In fact, we only could find four genres of literature: each one for each age.
Studying the times in which each function of language appears, we will find four differentiated segments: Stereotype, Illusion, Pragmatism, Justification. As was exposed in [4] by Vladimir Propp, the Folk Tale is composed of a strictly ordered sequence where some segments might not have to appear. After studying his morphology, the notation was reduced to write a story like in the pentagram of a score, counting a sequence of 4 by 4.
After studying The Hero with the Thousand Faces [5] of Joseph Campbell, the same segments in a different order can be found. So a more complete theory could be:
Chanzas genre. EJIP. The baby is surprised by fabulous and jokes. This will help him to imagine alternative worlds in order to form the basis of his own philosophy.
Folk tale genre. EIPJ. The child is told stories about heroes who break established morality. This type of story will help the child understand how horizontal relationships are generated with other peers, such as marriage.
Heroic genre. EJPI. The teenager is told stories of figures that came out of the nest. This kind of story will help the adolescent break the rules and seek his or her own legacy.
Thriller genre. EIJP. The adult hears stories of complex behaviors in the tribe. Through this literature, the adult will develop a consciousness that can make him or her a leader.
In conclusion, until we can put our algorithms on our knees and tell them a fairy tale, I doubt that they know how to situate themselves within the history of humanity.
References
[1] Dylan F. Glas, Takayuki Kanda, Hiroshi Ishiguro, Norihiro Hagita, (2012) ”Temporal Awareness in Teleoperation of Conversational Robots”, IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and Humans, Vol. 42, №4, pp.905–919
[2] Hiroshi Ishiguro (2006), “Android Science”. Scientific American. 294 (5): 32–34.
[3] Falk, D. (2004). “Prelinguistic Evolution in Early Hominins: Whence Motherese?” Behavioral Brain Science, 27, 491–503; Discussion 503–583.
[4] Vladimir Propp (1984), “Morphology of the Folk Tale”
[5] Joseph Campbell (1993). The Hero with a Thousand Faces. London: Fontana Press. ISBN 978–0–586–08571–4.
