MACHINE LEARNING AND COGNITIVE COMPUTING
Machine Learning and Cognitive Computing
Machine Learning and Cognitive Computing
Technological evolution has been one field of wonder. Everything currently done by humans can be done similarly or even better by use of artificially intelligent machines such as expert systems and decision support system. In improving technology, scientists have already proved that there is no more time to sleep. Most people, even to presents still know that a newly bought computer must first be installed for application programs before one starts to use them but this is currently fading away. The scientists are somehow tired of writing application programs; this has been revealed through the introduction of learning systems. With all these inventions, scholars have given their thoughts in terms of their future expectation of the current technology, disadvantages, and advantages of this technology. Therefore, this research will concentrate on the social implications of machine learning and cognitive computing.
Background of Machine Learning and Cognitive Computing
Machine learning is a type of artificial intelligence in which computers can carry out their obliged duties without necessarily being programmed, but the machines use pattern recognition to find hidden insight from analysis of data fed into them. In 1957, Frank Rosenblatt worked out a neural network for computers which simulated the human brain, and this was a milestone in making learning machines. However, this was preceded by various inventions like Turing test by Alan Turing which aimed at testing the computer’s ability to act like a human being in 1950; computer learning program by Arthur Samuel in the game of checkers where the computer improved after every game it played; and, using the computer’s game patterns in making a program.
The attempt to make a computer in a way that it can mimic the human reasoning and brain model is also known as cognitive computing. In self- learning, the computer will use systems such as natural language processors, data miners, and pattern recognizers. This means that in the near future, machines will take over every duty that requires expertise and make you and your colleagues jobless.
These technologies are employed amidst us although most people have not been in positions to know them; for those who have already identified them, have reacted to them in various ways. One of the common areas where this technology has been employed is in Facebook where news feed for friends you always communicate with appear first when you log in and those whom you rarely say hi to come later. The Facebook system has, in this case, learnt from the experience on your Facebook activities.
One of the major concerns of the general public on these systems is that they may soon take over the industry and render the current employees jobless according to Ford (2013). According to studies, however, it is clear that these insecure employees have hopes in working with these systems. The machines take data from experts, and for these employees to be able to work with these machines, they have to attain a certain level of expertise which is possible because the employees get to experience day by day in their working environment.
The next concern to the public on these systems, according to Rogers and McClelland (2014), is in knowledge acquisition. The machines do not have common sense, and this would imply that the machine lacks the ability to select the type of knowledge it should acquire; being either good or bad. This may lead to inheritance of bad characters from the experts or the user which are not acceptable by the general public. This is also connected to the fact that these machines cannot learn from their mistakes. The other concern related to lack of common sense in these systems is the inability to give proper justification for their actions in cases where creativity.
The other concern of the general public on these systems is the flexibility of the domain in which these systems will be working. Most people think that the domains of these systems are so narrow that they can never employ their use. However, this is not true because from the definition of what these systems are, they have the capability of working irrespective of the area they are employed. They have knowledge bases and are not limited in their application[ CITATION Fla12 l 1033 ].
Computer systems having the ability to think and act as human beings means that they can work together to make themselves even better than the way human beings make them; this implies the end of human race as per Bostrom (2005). There is a possibility that these systems with time will have the ability to regenerate themselves and render human race as a threat.
The other concern of learning machines is in the definition in different contexts. This was arrived at since meanings of languages always depend on the settings as well as the behavior of the speaker in that particular setting, but it is not possible to expose the learning machines to all these settings to have it record interpretation rules relevant to each of these settings[ CITATION Kra12 l 1033 ]. This will imply misuse of language in different settings, and this is a moral threat.
Most religions have also opposed the development of robots and other intelligent and learning machines arguing that human beings are trying to compete with their gods. With this mind and perception, they have failed to promote the development of these machines as well as even discouraging their uses because they fear that these will render religion obsolete as Aguilar, Santamaria-Bonfil, Froese, and Gershenson (2014) argue.
The final concern on learning and cognitive computers will be their ease of operation by the physically challenged. This is a group who will employ the use of sign languages especially the deaf and the dumb. It is very difficult to incorporate sign language on the memory of machines and also to communicate. This will mean that such people will be left out of this technology despite their high level of expertise and education but we cannot conclude this because innovation is beyond limit[ CITATION Jia14 l 1033 ].
If cognitive computing and machine learning technologies were advanced and available today, every activity which requires expertise would be done using machines making human labor very little. Some of the potential future of these technologies include internet of things (IoT), cure of hazardous diseases like cancer and big data processing.
Internet of things (IoT) is an artificial intelligence technology in which objects are networked and have the ability to send and receive data without human intervention. Such things include homes, vehicles, and wares. In medicine, cognitive computing is optimistic ion finding a means of curing cancer and removal of tumors without physical surgery. Big data is the important enormous data got from every day’s business undertakings. Cognitive computing and learning systems will have the ability to find patterns in this data that they can be useful in day to day operation of business because it is too large.
One of the reasons why I support the continuation of this research is because it will improve most of our daily undertakings; for instance, it will improve transportation where there will be no human drivers who are prone to human mistakes in driving hence reducing the rate of accidents. In health care facility, these technologies will improve research and diagnosis for dangerous diseases like cancer. This will reduce premature deaths caused by these diseases. In the next five years, I do not expect much improvement on these technologies because most people are not still aware and this will take time. However, I expect to see driverless public service vehicles in our streets and smart homes.
Almost all serious companies have adopted the use of artificial intelligence in one of their production process of the other. Despite employment of this technology, there are numerous gaps that the use of machines cannot fill without the intervention of human beings. This means that human beings together with the machines will be able to work together so that they can perform perfectly. It is clear that machines are reliable and never get tired, they are also not affected by mood swings and can never be persuaded. These might be advantages of machines, but these don not qualify them as perfect replacement of human labor.
Aguilar, W., Santamaria-Bonfil, G., Froese, T., & Gershenson, C. (2014). The past, present, and future of artificial life. Frontiers in Robotics and AI, 1, 8.
Bostrom, N. (2005). In defense of posthuman dignity. Bioethics, 19(3), 202-214.
Flavin , P. G., & Totton, K. A. (2012). Computer aided decision support in telecommunications. New York: Springer Science & Business Media.
Ford, M. (2013). Cloud artificial intelligence create an unemployment crisis? Communication of the ACM, 56(7), 37-39.
Jia, S., Dong, Z., Li, X., Pang, X., Guo, B., Wang, S., & Wang, K. (2014). Distributed intelligent assistance robotic system with sensor networks based on robot technology middleware. International Journal of Distributed Sensor Networks, 10(6), 908260.
Krause, P., & Clark, D. (2012). Representing uncertain knowledge: An artificial intelligence approach. New York: Springer Science & Business Media.
Rogers, T. T., & McClelland, J. L. (2014). Parallel distributed processing at 25: Further explorations in the microstructure of cognition. Cognitive Science, 38(6), 1024-1077.