Maria from Metropolis
This opens a truly fantastic vista of exploration and high adventure...
William Grey Walter
I've never thought much about robots. As a young boy I‘ve read Asimov and other science fiction books in which automata, androids and clever machines were the main characters. But this was merely due to the fact that I have always been a booklover that enjoys reading almost anything. Actually I am not even a big enthusiast of the genre. Until recently I also didn't pay much attention to the sparse robotic appearances on TV and other media. The only image I can recall from my youth it is that of Maria, the female robot in the film Metropolis of Fritz Lang. Not so much for being a robot but because I liked the idea of a metallic woman.
My debut in the universe of robotics was the result of personal evolution. By the end of the nineties I’ve plunged into an experimental technocreative course using successively digital, internet and algorithmic art, artificial life and robotics. Being an artist I felt the need of projecting most creative concepts into real space. Robots appear to be a good solution for implementing emergent behavior in the physical world.
Since I first created a robot in 2001 (in fact a robotic arm) I didn't stop anymore. Building autonomous robotic entities is equally exciting as an artistic creation or as a scientific discovery. Actually, it is a combination of both. Hence I have learned a lot about these peculiar machines, to the point that they are now my daily mates in several projects.
Humanity's interest on robots is obviously much older than mine. To illustrate this it is worth to take a quick historical trip.
The Artificial Man is a creature present in many ancient mythologies. Made out of body or corpse parts, hybrids, animation of inanimate matter or phantasmagoric elements, a plethora of specimens have aroused human imagination throughout history. The Golem, an artificial humanoid made out of mud, is a well-known character since early Judaism, although its most famous version made its appearance in the 16th century Prague. With a more scientific and rational approach, the ancient Greeks had already the idea of building machines able to operate on their own. It is well known that Aristotle said that "if every tool could perform its own work when ordered, or by seeing what to do in advance, master-craftsmen would have no need of assistants and masters no need of slaves". Combining mathematics with mechanics, the Greeks have achieved the first inventions of the kind. Around 300 BCE Archytas of Tarentum have allegedly created the first robot, an automaton in the form of a wooden bird, able to simulate flight. Then we need to wait for the 13th century to record another important milestone: "The Book of Knowledge of Ingenious Mechanical Devices", also known as Automaton, written by the Arabian Al-Jazari, where many mechanical inventions were explained by text and construction plans. It is probable that Leonardo da Vinci has read this work, as expected from Arabic influence on the Italian Renaissance. Anyway he developed many mechanical devices, from airplanes to a helicopter and a self-propelling vehicle predecessor of today's cars. The drawings to build a humanoid robot also survived to the present days.
In the 18th century appeared the first wave of machines able to simulate human and animal behavior on the grounds of countless automata developed by famous clockmakers and other craftsmen. These robots used to amaze nobles and rich bourgeois by their ability to play music, draw, write or engage in chess games. Jacques de Vaucanson digestion duck is one of the most admirable of these. The duck flaps its wings, eats, digests grain and defecates. Vaucanson also invented the first punch cards for the automation of the weaving process.
These historical examples are clearly based only on a more or less complex mechanics. They lack true autonomy and miss that lively spark that we recognize in animated organisms. In this context Frankenstein remains the most dramatic attempt to conceptualize the autopoiesis process, defined by Maturana and Varela as the self-creation that characterizes living systems.
In the 20th century an extraordinary acceleration in machine evolution occurs. The computer made viable large scale work on artificial intelligence and freewill machines. The computer allows the generation, testing and development of a new kind of intelligence which, even though being artificial, maintains its "intelligent" quality.
How to simulate complexity in machines became the main question. In the forties the debate between analog and digital came to light. Although the matter was never consensual, it is now clear that the first phase of artificial intelligence, relying chiefly on hard digital computation and human intelligence models, has completely missed the point. Anyway, after decades of experimentation, by the end of the nineties, it was made public that a computer had gained a chess game against the world chess champion Garry Kasparov. The event was presented as the triumph of the machine superior intelligence. The IBM computer Deep Blue established a milestone in the proclaimed machine revolution. However, this extraordinary processing capacity did not solve some basic questions when considering the autonomy of artificial organisms. Although Deep Blue had an "idea" of the chess board it needed an assistant to move the pieces.
A parallel tendency relying on apparently less ambitious experiments won relevance, in what concerns a new kind of machines that stem from the ideas of MIT teacher and father of cybernetics Norbert Wiener.
The neurophysiologist William Grey Walter created back in the forties a set of "Machina Speculatrix" aiming to develop such attributes as: "exploration, curiosity, free-will in the sense of unpredictability, goal-seeking, self-regulation, avoidance of dilemmas, foresight, memory, learning, forgetting, association of ideas, form recognition, and the elements of social accommodation".
Instead of assuming the human being as the chief model, Walter tried to engender small robots with emergent behavior. The "Turtles", as they were coined because of their shape and slow speed, performed complex actions based on simple rules. By using sensors they avoided obstacles, responded to light stimulus (phototaxis) and find their way to a recharging location when their battery power was low. Walter used mainly analogue electronics as opposed to his contemporaries Alan Turing and John Von Neumann who engage to promoting digital computation.
By the end of the eighties Rodney Brooks, influenced by the work of Grey Walter, advanced his bioinspired robots supported on a Subsumption Architecture, depicted as the decomposition of intelligent behavior into a series ranging from simple to complex layers. Each layer subsumes a lower layer, increasing complexity accordingly. Concepts like embodiment, situatedness and dynamicism are also crucial to this behavior based robotics, whose agents in the first experiments looked like insects.
Not less relevant for this brief history are the concepts of emergence and out of control. The study of several natural phenomena proves that complexity is the result of very simple, basic and local interactions. The Whole is in fact often superior to the sum of the parts, as put forward by the Gestalt theory. Also the principle of self-organization, in matter and in living organisms, suggests that promoting bottom-up strategies is more important than to build top-down structures. Systems based on simple rules and multiple free-willing agents tend to be more robust than non-cooperative organizations.
The robotics explosion that occurred since the last decades of the 20th century follows parallel paths. While bioinspired robotics wages on biological mechanisms and out of control approaches, the old line based on an intense machine control survives mainly in industrial and humanoid robots, even though a combination of the two approaches is also frequent. Also, analog and digital are now seen as partners more than antagonists.
Human fascination for robots has however very little to do with this kind of technical and conceptual debate. It is their lively behavior that we find so attractive as we tend to identify autonomous movement with life itself. There is actually a genetic explanation for this phenomenon. Self preservation recommends that we immediately relate independent movement with being alive. If the stone moves it can be a crocodile.
On the other hand, at the same time that we consider ourselves as a hypothetic and illusive top of the tree of life - much above all the other life forms -, we also feel an enormous solitude that stems exactly from the same fact. There is an emotional resemblance between our present quest to encounter an extraterrestrial life and the will to create machines more advanced than us. To pursue those objectives, we send messages to the universe and build robots that are each time more sophisticate and smart. Given the evident risks of both endeavors, any rational explanation to these purposes is out of the question. Human loneliness may be the reason.
Although robotics is only biological through its bioinspiration, it simulates life and raises many questions related to its definition and manipulation. There is nothing that requires future life forms to be built just out of cells and living tissues. Silicon, metal and plastic can perfectly be used as well. And probably the most practical will be a combination between bio and bot, mechanical and living parts interacting with each other. Donna Haraway's cyborg concept, hybrid of flesh and machine, is the precursor paradigm of such an approach.
Robotics must be seen as an amplification of life: a new kind of species born from a postnature context in order to stimulate, expand and enhance the available intelligence on the planet.
<- Main
contact
