The Current State of Art in Artificial Intelligence
The Current State of Art in Artificial Intelligence
As from the creation of artificial intelligence (AL), the field has undergone diverse changes with regards to its coverage. It has transformed from being an idealistic concept to a pragmatic discipline. There are a number of significant breakthroughs that have resulted to attainment of the long desired artificial intelligence. This essay looks at the current state of art in artificial intelligence. It starts by highlighting then inception of artificial intelligence, the process that it went through to facilitate its transformation to become pragmatic and an insight into its historical development. The essay will end by looking at the possible future of AI.
To fully understand the developments that artificial intelligence has gone through over the years, it is essential to briefly look at the historical context of its origin. The formalization of machine intelligence has been in the interest of humans for as long as the human history exists. However, efforts to launch the initiative were futile until during the Second World War that presented a favorable environment for establishment of initiatives to facilitate expansion and formalization of machine intelligence. The need to initiate machine intelligence came about at a time when designing machines with the ability to respond to external feedback emerged as a vital requirement. During WWII, an automatic flask control system was developed by a United State’s professor called Norbert Weiner. The flask control system was capable of adjusting itself in relation to the input and the feedback that was received from a radar. The radar and its input and feedback system was essential in understanding adaptable systems. Weiner’s perception of what is generally understood as intelligent act is that it results from feedback mechanisms. Weiner’s concept gave rise to the establishment of the feedback theory.
The early attempts towards development of artificial intelligence had some points of similarity with the current AI. However, it is generally believed that there is need for digital computers to facilitate the advancement of AI. It is due to this reason that several efforts were made to ensure that digital computing system was made available to support the functioning of artificial intelligence. Artificial intelligence was launched as a branch of computer science during an AI conference that was convened in 1956. The term ‘Artificial intelligence’ was introduced in 1955 by McCarthy as the origin of the concept is attributed to Herbert Simon and Marvin Minsky among others.
Several developments have been noted in the field of artificial intelligence starting with the development of Lisp programming language. This milestone was reached in 1958 and it provided an essential pillar of the community of artificial intelligence. The programme still exists in different forms. The year 1968 also saw a great development in the history of Artificial intelligence that came about as a result of the establishment of natural language processing program SHRDLU that was able to manipulate virtual geometrical objects through processing and following English written commands. The earliest developmental phases of Artificial intelligence was highly linked to the academic community with the advocators for the system showing much optimism that with adequate support, the n realization of the system would take a short time frame. The initiatives that were launched to aid in creating AI were aimed at creation of an overall intelligence that is much similar to that of human beings. Inability to realize the overambitious objectives of the establishment of AI in the initial phase, known as the first AI winter posed a challenge to realization of the project as government funding was cut shot with the corporate bodies threatening to withdraw their funding. However, it is vital to look at the period when looking at the current state of art of artificial intelligence as it marked the launch and the initial innovations that formed the basis for the current stage.
There was another essential division of artificial intelligence known as connectionism that also faced challenges during the first AI winter of development of artificial intelligence. The subcategory concerned the attainment of artificial intelligence through distributed networks of nodes and links. Connectionism was believed to have the ability to offer solutions to several problems that AI faced including facilitating better understanding of natural language. However, the connectionism approached battled with several challenges that posed a challenge for it to achieve the desired objectives.
The first face was characterized by several obstacles hence making it difficult for significant progress to be realized. The field of artificial intelligence realized appreciable progress after the emergence of expert systems that were characterized by programs that are well adopted to provide expert knowledge and to offer solutions to problems in a particular domain. For example, improvements that were realized with the innovation of expert systems resulted in development of healthcare provision by facilitating provision of diagnoses. Developments continued to be observed in artificial intelligence with the second rise of connectionism. At the stage, several problems that were initially encountered with the connectionism technique were brought to an end. For example, the development solved the problem relating to the difficulties in approximating nonlinear functions by use of algorithms like backpropagation.
The phase also faced its share of challenges that came about as a result of failure to meet the desired results as was expected. Financial problems emerged and threatened to fail the attainment of artificial intelligence but it was evident that a new trend was starting. The next trend that hinged the development of artificial intelligence towards its contemporary state was referred to as the phase of commercialization and practicalization of techniques and research topics. The initial goals that were attached to the development of general intelligence slowly reduced as efforts that were made to facilitate the attainment of the desired results failed to bear fruit.
Currently, artificial intelligence has reached a point of permanently affirming its place among other disciplines after completion of excessive optimism. It has also undergone great transformations that have enabled it change from an idealistic concept to a pragmatic discipline. Currently, Artificial intelligence has been adopted by the academic community through establishment of a significant toolbox for use by the IT industry. Today, AI is used in several spheres of human life as people have heightened their level of use of electronic gadgets in the current information era. Virtually all technological gadgets and software are capable of learning including mobile phones and cars. It is interesting to note that the developments in artificial intelligence have reached a point where people can rely on technological appliances and software to help them perform tasks that the human brain can hardly execute. Many people rely on Google that has the capacity to understand peoples’ interests so as to offer them more effective search results. Better still, with the emergence of hyper-cheap electronics and robotics; it is clear that humans will use several intelligent machines to perform a wide range of tasks. The machines have undergone gradual developments to the extent that they can currently be depended on for execution of complex tasks that the human brain can hardly perform.
There are three essential breakthroughs that are credited for the current status of Artificial intelligence. One of the developments is the cheap parallel computation that enabled the typical computer processor to perform different processes simultaneously. The addition of specialized parallel computing chips to supplement the PC motherboard facilitated multiple processing that enable neural networks to operate in parallel. The invention gave room for adoption of neural networks that can incorporate multiple connections between their nodes. Currently, cloud enabled companies apply neural nets that run on GPUs to execute complex tasks within limited time. It is evident in the way Facebook currently uses the system in identification of subscribers’ friends. The second development has to do with the availability of an appreciable avalanche of data that is gathered about the world. The data provide training that is essential for effective functioning of AIs. Availability of the big data like web cookies, self-tracking and terabytes of storage among others offer schooling to facilitate effectiveness of AI. Artificial Intelligence relies on the large data set as it functions in a similar way as the human brain that needs to be subjected to repetitive examples to grasp and retain concepts. Another essential development that characterizes the current status of AI is the availability of better algorithms. It led to the adoption of deep-learning algorithms that were ported to GPUs. Although deep learning code cannot effectively generate complex logical thought, it is a vital component of today’s AIs with Google search engines and IBM’s Watson inclusive.
The developments that have been witnessed have facilitated the performance of given forms of computer visions and natural language processing. Currently, it is possible to convert good-quality printed text to electronic text without making many errors. The ability to have the printed texts converted to electronic text is considered as a great advancement although old documents and the documents with low qualities still present some challenges during conversion. Developments in artificial intelligence can even be observed in the current generation of ATMs that have the ability to read hand written checks. There are tasks in which computers are realized to perform much better than human beings. For example, the machines perform analysis of some types of technical images more effective than humans can. Web search engines are also credited for their ability to find specific documents with remarkable levels of accuracy and at fast speed. Students of language do not have to rely on their hard copy dictionaries for translations as then goggle translator offers them good translations that can help them in the learning process except for some funny errors.
A part from the developments discussed above, there are several possible developments and designs that are likely to be coming in the next decade. Several initiatives are underway to facilitate creation of ‘hard’ intelligence. The endeavors are no longer idealistic as the current developments that have been realized and the developmental capacity of the advancements have had people convinced that they can help to move the world to greater heights. The initiatives have gained value and support as they are regarded as practical unlike the early stages when efforts to realize artificial intelligence were perceived as utopian and mainly idealistic. For example, the OpenCog initiative that is based in Hong Kong is promising to result to significant developments in the field of AI.
Research into AI should be directed in three directions to facilitate future advancements in the field. It is necessary to step up initiatives aimed at creating a particular framework composed of different data structures, algorithms and techniques that function on the basis of some knowledge with close consideration of the environment. The initiative is already launched by OpenCog and other players should join to make it more vibrant. The second direction can focus on how humans can serve as simulators of the environment to give room for the occurrence of vital evolutions. Besides, it is necessary to implement the third direction that involves direct simulation of the human brain.
There are other opportunities for attainment of artificial intelligence. For example, future application of a powerful general intelligence machine can help in figuring out the procedure that can be taken to perform AI tasks. As technological advancements increase day by day and with the observed improvements in neuroscience, it worth projecting that an emulation of the human brain can be built in the future. Advancements in the emulation to have it few magnitudes faster than the human brain can lead to attainment of a super-brain. It can only be achieved if neuroscience is advanced to another level that takes keen interest in advancing the current developments.
In conclusion, based on this discussion, it is clear artificial intelligence has undergone several developments to reach its current stage. The development was characterized by several challenges that resulted from the overambitious goals that were set in the initial face and that threatened to stop further developments. Governments and other financiers threatened to withdraw their support but later achievements motivated them to back up developmental initiatives. As such, artificial intelligence has reached a stage where humans heavily rely on it to perform several tasks. It is evident in the way humans use technological gadgets like computers, mobile phones and ATM machines to perform technical functions.