The advancement in technology enabled man to create various types of intelligent robots that are now widely used by humans. These technologies, such as the evolutionary robotics, coupled with the need for people to automate daily and recurring tasks brought about the idea of creating autonomous robots. Today, intelligent robots are being beneficial in the performance of different tasks as science discovered ways to make these human creations become dependable and interactive. Needless to say, there is a shift from the previous study and design of robots from strict industrial purposes, into human friendly intents. Evolutionary robotics foresaw the development of robots that are intelligent enough to co-exist with human. However, robotic intelligence can never be equal to that of the cognitive ability of humans as the human brain and cognition is too multifaceted to permit replication. The study of evolutionary robotics cannot be associated with the study of human cognition because human intelligence is a complex structure that cannot be duplicated through the use of research and studies on robotic intelligence.
One reason why Evolutionary robotics cannot be associated with the study of human cognition is that, despite the advances in technological science, there is no way that man can be able to simulate the human brain and its complexities. Scholars who are strong proponents of the creation of cognitive robots are supportive of the notion that these human masterpieces will soon have the ability of perception, attention, anticipation, planning, memory, learning, and reasoning just like humans. However, notwithstanding the assertion about the possibility of human like capability of robots, it is proven that the robots that are currently common today do not have the capacity to learn, and though they have memory, it is far from duplicating the memory of a human brain. Accordingly, “at this point most existing robots are used in industry (think assembly lines), and most of them don't even have any perceptual abilities at all; they are programmed to do one thing, and one thing only” (Cognitive). There can never be a robot that can think and perceived things like humans do. Even evolutionary robotics that has gained its popularity during the 1990s failed to prove the replication of human like intelligence in a machine.
Proponents of evolutionary robotics, however, reasoned that the motivation of their study is based on the “understanding of cognition in its most basic sense, and sees ER as a useful testbed, a methodology for generating synthetic or artificial agents in relatively prejudice-free fashion that can then be investigated and analysed” (Harver et al). Evolutionary robotics are based on the Dynamical System, which, when used as an approach to cognition, it implies a commitment to explaining cognitive phenomena in the mathematical language of DS” (Harvey et al). According to scholars, while the synthetic nervous system bears a minimal resemblance to that of the human nervous system or any living organism for that matter, a successful experiment is significant enough as it became a form of Existence Proof. Evolutionary robotics allows for the “exploration of the minimal conditions for a given behavioural capability” (Harvey et al). In addition to these arguments, the Dynamical system is characteristically counter intuitive, so that the anticipated behaviour is as shown, with the sensorimotor interactions with its environment, and as noted this particular process is difficult to predict or design for. The argument further went as to describe the natural living structures being intended as a dynamical system through the Darwinian evolution, so that the artificial equivalent of the development can be correlated with the process.
However, the argument on the basic understanding of evolutionary robotics and other artificial intelligence cannot be associated with human cognition due to the fact that robots cannot be used to represent human intelligence. As stated by Brooks, “I and others believe that
human level intelligence is too complex and little understood to be correctly decomposed into the subpieces at the moment” (Brooks).
Robots or artificial intelligence cannot manage appropriately and in a timely manner with the diverse modifications that may happen in its active surroundings. When a robot is put into an autonomous environment and tested on its capabilities, it is likely to fail in areas that it is not programmed to do. In addition to that, it does not have the reasoning capacity like that of humans to decide beyond that which is programmed to do. Again Brooks maintained that, “when researchers working on a particular module get to choose both the inputs and the outputs that specify the module requirements”, he believed that there is minimal opportunity that the work they do will fit into a wholly intelligent system. Accordingly, there is the chance that researchers, due to pressure may circumvent a particular case of study to arrive at a decided conclusion.
Those who advocated the capability of robots and AI’s to deal with the environment cited how the robots of today were able to adapt through learning by its instincts. These robots are pre-designed in a way that they can perform learning at levels comparable to that of a minimum intelligence level. For instance, these machines can be programmed to know to avoid objects, recognize their master, follow directions, and locate objects among others (Brooks). The MIT Artificial Intelligence Laboratory has created a robot that has infrared proximity that is sensitive enough to avoid objects. It was a magnificent creation that it can be able to grasp objects and determine their weight. This creation was made many years ago, at one can only marvel at how much the technological advances in robotics had reached at this time. Brooks then expressed that time will tell what their creations will prove in the real world; and robots today have proven to be more adept than they were many years ago.
Admittedly, the robot creations today are “skilled” enough to be of help to other people. For example, a dishwashing robot was created in Japan; it was smart enough to handle the most precious ware as it was created with intelligent sensors that make it handle its tasks with efficiency. There are other exemplification of the successes of robots and artificial intelligence, proving how technology was able to capture and duplicate the human intelligence. “Robots go beyond telemanipulators in that their sensorimotor capacities enable them to deal with a greater variety of circumstances and events without on-line human involvement” (Haselager). But despite these facts, these are not enough foundation to base the study of cognition through evolutionary robotics. For one, the simulation approach of evolutionary robotics has been highly debated since the inception of the ER. It is to be noted that the evolution being presented by ER is a very simplified duplication of the natural evolution as presented by the Darwinian Theory. This lends a lesser value to the ER approach, as natural evolution considers all the biological interactions, and the causal relationships of the biological systems, in contrast to that of the ER approach that relies merely on the hypothesis during the experiments.
The robots we have today may have the autonomy to do the things that they are purposely made to do, but still, it remains that they do their tasked as set by their programmer. There may be a reduction of human involvement in the functioning of the machine, making it “act independently”, but the great amount of time and human involvement in its creation does not make it less a robot. In short, no matter how much of an intelligence in given to a robot, it can never function as a human being. A creature that is made through artificial means lacks the characteristics that can be associated with a creature that has undergone the millions of years of evolution.
The process of evolution in itself warrants the arguments against the use of evolutionary robotics in cognitive science. “The point about creatures is that millions of years of evolution have equipped them with a fierce determination to survive” (Malcom). This is how the survival of the fittest of the Darwinian evolution was founded. Evolutionary robotics, on the other hand, is defined as the, “attempt to develop robots and their sensorimotor control systems through an automatic design process involving artificial evolution” (Haselager). Moreover, the ER disrupts the natural evolution process of Darwinian evolution as the designer pre-arranged the course of evolution, choosing the inputs and even the outputs of the process. Unnatural interventions are made by setting the experiment thus weakening the results of the study, “some additional intervention is to needed to canalize the evolutionary process into the right direction” (qted in Haselager). Another argument against the use of ER is that fact that what usually went through the transformation is the control system and not the physical or morphological characteristics of the robot.
One of the most used robots is the khepera, a small pre-made robot, with
specific sensori-motor capacities that can be controlled through neutral
networks. In the context of artificial evolution, most often khepera
simulators are used instead of the real robots.() It is important to
realize that during the evolutionary process, the khepera itself does not
udergo any changes at all. There is no equivalent for this in evolution (Haselager).
Based on the context above, there is reason to question the validity of evolution that changes only the control system, but does not introduce any changes in the physical body. This fact served to be a limitation of the ER compared to the study of natural evolution; that is, human interventions are needed to refine the results of ER.
Robots do not embody the characteristics of living organisms. Living things have the capacity to acquire the materials they needed for energy and survival. For instance, people eat and through the process of metabolism the digestive system converts the food into nutrients that are needed by the body. Each living organism has also the capacity to stay in balance with its environment. These basic attributes cannot be observed in even the most recent discoveries among the robots today, though one may argue that modern robots have the capacity to regularly check their level of energy and would take the required action in cases when its energy is going low. Just like humans and other living organism, failure of the machine to self-check and maintain its body requirements may result to damaging effects leading to malfunction or even total damage. However, these are mere functional characteristics, as one can always turn off the robot for a considerable time, and still may function normally when turned on. This would be impossible for any living organisms as each of their parts is needed for continued existence.
Conclusion
The advancement in technology today made people create robots that can be used to study humans such as the use of evolutionary robotics in the study of cognition. Though there are scientists who supports these new developments, there remains to be a question on the validity of the association between ER and cognition. For one, the human brain is too complex to warrant its simulation through controlled robotics. In addition to that, robots do not have the capacity to survive in changing environments, a far contrast to living organism, especially to humans that have the cognitive ability to adapt to diverse changes. There are just many basic attributes that make robots and humans different, thus, making the results of the research on ER and the study of human cognition of lesser connotation.
References
Brooks, R., Intelligence without Representation. MIT Artificial Intelligence Laboratory.
Cognitive Robotics. What is Cognitive Robotics? Retrieved from www.cogsci.rpi.edu
Harvey, I., Di Paolo E., Tuci, E., Wood, R., Quinn, M., Evolutionary Robotics: A New Scientific Tool for Studying Cognition. Centre for Computational Neuroscience and Robotics
Hasealger, W., Robotics, Philosophy and the Problems of Autonomy. Retrieved from ww.socsci.ru.nl.
Malcom, C., Why Robots Won’t Rule. Retrieved from www.dai.ed.ac.uk
