The best intelligence known in existence resides within the human nervous system. Artificial Intelligence is defined as “the study of how to make machines behave intelligently, to solve problems and achieve goals in the kinds of complex situations in which humans require intelligence to achieve goals (Fetzer).” People have conducted studies of Artificial Intelligence from numerous points of view. Steady progress continues in this field, but humans have yet to understand human problem solving or develop accurate theories of higher mental function. Conversely, Artificial Intelligence does exist currently with an ability to complete various functions. While artificial intelligence can perform a multitude of tasks quickly and efficiently, it cannot comprehend a deeper knowledge or fully understand the world.
History of Artificial Intelligence
Ever since the seventeenth century, people have studied human and animal neurophysiology and neuroanatomy to try to build a sufficiently similar intelligence. This dream began with Gottfried Wilhelm Leibniz, who believed in reducing thinking into calculation. During his time, he helped to invent the mechanical calculator and developed calculus independently of Isaac Newton. Leibniz had a vision of “Mathesis Universalis” which, “was a vision of universal science made possible by a mathematical language more precise than natural languages (Larson, 2016).” This project never finished, but it helped to design modern symbolic logic. (Larson, 2016)
Save your time!
We can take care of your essay
- Proper editing and formatting
- Free revision, title page, and bibliography
- Flexible prices and money-back guarantee
Along with these precise calculations, people have studied human intelligence on a psychological level through experiments where human subjects solve problems. During the Enlightenment Era in the eighteenth century, a philosopher and proto-psychologist, Étienne Bonnot de Condillac envisioned humans with what he coined, “the inward organization”. He wondered what happens to information when it enters a person. Condillac drew inspiration for his thinking from a previous philosopher, Thomas Hobbes, who has argued that thinking was precise calculation. (Fetzer) (Larson, 2016)
The first true steps toward Artificial Intelligence as a technology happened in the nineteenth century by Charles Babbage. Babbage created an engineering project aimed at constructing a machine able to think. His concept for his “Analytic Engine” incorporated components of modern day computers. Unfortunately, the British Association for the Advancement of Science denied additional funding to the project and the project dissipated. (Larson, 2016)
In 1955, John McCarthy coined the phrase “Artificial Intelligence”. By 1956, Artificial Intelligence launched as an independent research field. The field met enthusiasm from scientist everywhere. The major goal was to answer the question set forth by code-breaker, Alan Turing, in his paper which began, “I propose to consider the question ‘Can machines think?’” Many challenges were faced throughout the research process with differing amounts of difficulty. These forty years of research laid a foundation for the breakthroughs to come in the near future. (Larson, 2016)Modern Artificial Intelligence and the Trajectory in the FutureThe swift development of the Internet in the 90’s allowed for the use of data on a large scale. Successful Artificial Intelligence today uses pages from the Internet to gather data in order to from that information into a proper response. In addition to data, the Internet created a demand for automated services to search and organize information from the expanding Internet. Modern Artificial Intelligence is driven by data and known as Empirical AI. The name Empirical AI derives from its roots in philosophical empiricism, “knowledge about the world is largely acquired, or learned, from experience (Larson, 2016).” This contrasts traditional AI which assumes, “that a significant part of human knowledge is not derived from experience but is “fixed” in advance in the capabilities of the brain or mind (Larson, 2016).”Because of the large number of obstacles faced against Artificial Intelligence, the pessimism naturally exists. In theory, Artificial Intelligence appears as fundamentally unlimited to human intelligence, yet the limits of Artificial Intelligence exist. Data, model complexity, and compute time all limit machine learning to a degree.The general consensus today is that Artificial Intelligence will never have a consciousness. What some perceive as limits, others see as obstacles to overcome. In reality, people do not have the scientific knowledge to know the limits of Artificial Intelligence.Function and Lack of Understanding
The classic AI, Deep Blue, notoriously is known for beating Grandmaster Garry Kasparov in 1997. The machine learned every possible chess board and could compute 200 million boards per second. The machine was successful because of the finite amount of possibilities. However, the machine did not have strategy or understanding of the game.
Artificial Intelligence commonly gets compared to human intelligence for obvious reasons. Noriko Arai wanted to compare these two directly by creating an Artificial Intelligence to take the entrance test to the University of Tokyo as a benchmark. Furthermore, the research project aimed to predict future outcomes in the job market. (Arai, 2017)The robot, Todai Robot, uses search tactics to scour the internet to find answers to questions by using keywords and proximity. The robot parses math problems into a language that it can process to produce an answer. For the essays, the AI takes sentences from sources, combined them, and optimized them to produce an essay better than most students. (Arai, 2017)Todai Robot did not pass the entrance exam, but did score higher than 80% of students. The concerning part is the simplicity of questions that it can get incorrect.
This question in the English section of the exam is uncomplicated to any English speaker. After learning 15 billion English sentences using deep learning technology, Todai Robot could not correctly answer this basic question. Noriko Arai frequently states that the Artificial Intelligence does not have the ability to understand like humans can. It can write an entire essay without actually having an understanding of what it means. (Arai, 2017)
Deep LearningHistorically, games such as Chess and Go were popular for Artificial Intelligence because of their traditional, less complex nature. Elon Musk’s company, OpenAI, set out to beat the best players in “Dota 2”. This “MOBA” (multiplayer online battle arena) is one of the most complex types of games available today. Because of this, the game had to be reduced to a one-on-one on one character only, but the number of possibilities is still unfeasible and highly complex. The had the AI play against itself with increased speed in multiple games at the same time. It started with random movements, but became more efficient with each game. In only two weeks, the AI was as good as the best players of the game. Eventually, it developed strategies that require a high level of thinking to achieve, only to counter those strategies after a few more hours of practice. At that point, it was nearly unbeatable.
OpenAI hopes to create a full team of five bots across all characters in the game, but many remain skeptical. Interacting with others and knowing the interactions between over a hundred characters seems achievable to some. The processing power to obtain this goal appears unreal for modern times. Theoretically, this dream is only a larger scale version with more variables than the previous achievement. This compares to the way that solving chess is the same as tic-tac-toe but with more variables.
Conclusion
Throughout the history of Artificial Intelligence, the quantity of obstacles remains numerous, but human aspirations remain with the same vigor. Multiple methods have been designed to achieve the goal of Artificial Intelligence from different angles. In modern times, Artificial Intelligence does not possess the ability to understand deeply. While some may argue that this is a limit of Artificial Intelligence, it can also be viewed as another challenge to accomplish. Even if AI never reaches that point, it certainly will be used extensively for multiple purposes that it is needed for. It can perform tasks efficiently and learn quickly through deep learning.
Stuck on your essay?
