Joseph Lewers
ERH-102-06
Mrs. Smith
3-11-16
Artificial Intelligence
Advancements in the field of technology over the past couple of decades have redefined how we see our own future. Just in the past decades, we have developed computers capable of connecting people across the world, assisting in medical operations, and even carrying out military operations. After these advancements, the common belief is that in the future every aspect of our lives will have some sort of technological influence. To an extent, our generation has already come to that point. Almost everything we do involves a computer, whether it be driving to work, catching up with your friends on social media, even something as simple as making a cup of coffee, but our generation’s technology all still has one thing in common; it is controlled by humans. As of right now, computers are unable to think and make decisions for themselves. However, that may not be the case in the generations to come. Being developed right now is a technology called artificial intelligence where computers will soon be able to think for themselves and make decisions based on the information programmed into them. On a personal note, this new up-and-coming technology has a lot of relevance to me individually since I am pursuing two fields which are potential consumers of the technology. At the moment, I am studying mechanical engineering which, if this technology becomes a reality, will be a part of my everyday life as an engineer. I also plan on commissioning in the Navy, which is known for having the newest technology and could possibly try to weaponize it. Now it may seem like humans are on the path to a world similar to the one seen in the popular Terminator movies, but I would argue that those fears are irrational. There are practical uses for artificial intelligence in the medical field, industrial business, and education. But is this technology realistic and reliable enough to perform these jobs? And as silly as it sounds, should society trust the technology to not become self-aware and not obey what it is programmed to do?
Before even considering trusting technology with a human life, I had to learn the process of how this technology works. I read a number of articles to try and piece together the inner-workings that go into artificial intelligence to determine its trustworthiness. In the article Simulation-Based Evaluation of Learning Sequences for Instructional Technologies written by John E. McEneaney, he explains the base for all computed decision making starts with the person who writes the code. The person must program a hierarchy of values which dictates the morals of the system. For example in a system relating to medical uses, the programmer must tell the computer how to weigh survivability, success rate, and other factors before making its decision. Next I read the article Probabilistic Machine Learning and Artificial Intelligence by Zoubin Ghahramani which tells about another key component of artificial intelligence where computers could learn from experience. This is made possible by methods called probabilistic modeling and Bayesian optimization. In layman’s terms, probabilistic programming is simply when a computer goes through a list of possible outcomes to a situation and analyzes data it has recorded in the past. Bayesian optimization is the part that goes through all of the past situations and weighs the potential risks and benefits of each option according to the hierarchy of morals programmed into it and then chooses the best option. With these pieces put together, we could very well see the development of infinitely intelligent computers that continuously analyze situations and make decisions based on the moral guideline set in place by the programmer.
To me, just by hearing the concrete facts about what is being done to develop the technology, my question of whether this technology is realistic is answered. Since there is clear progress being made towards developing the software that would be used in developing artificial intelligence, we could see this technology being fully functional in the near future. These articles also reinforced my belief that there is nothing to be wary of if computers do become artificially intelligent. One fear that lurks in the back of society’s mind is if computers can think for themselves they might learn not obey commands they are given. These articles put those fears to rest by clearly stating how the computers have to go through a whole operation before making its decision and that the operation can only be set and altered by the programmer. Those are my two takeaways from the first two articles but they did not really touch on why the technology would be practical, or its reliability so I had to keep reading.
Then I came across the book The AI Business: Commercial Uses of Artificial Intelligence by P.H Winston and K.A Prendergast. The book goes into depth about the commercial uses for artificial intelligence in medicine, industry, and other electronics such as robots. The authors claim that that artificial intelligence will be paired with advance robotic technology to help in the medical and industrial fields (177). They predict robots could be used to assist or even perform medical procedures more precisely than humans which could significantly increase the survivability rate of high stakes procedures (179). They also see the possibility of robots being used in industrial settings. Robots and other intelligent systems could run 24/7 monitoring machinery and if something goes wrong, the system could diagnose and repair the problem (216). These are very relevant applications of artificial intelligence and when the technology does become a reality I believe these two fields will be some of the first fields it will be applied to.
Personally I was already aware of the practical applications of artificial intelligence from prior interest in the subject, but I do believe this book could have opened some people’s eyes to what this technology could be used for. If robots could assist with or perform high stake medical procedures flawlessly I think that would persuade patients of the technology’s practicality by eliminating human error which could save lives. Next, if robots or artificially intelligent monitoring systems were introduced to industries, factories could run 24/7 without having to hire workers to monitor the production. That would cut down the costs of running factories and essentially double production which without a doubt would persuade large companies of AI’s potential. If production rates of factories double it would not only be beneficial for that company, but the entire country’s GDP which would help our economy immensely and would most likely silence all doubts of about artificial intelligence’s practical usage.
Before doing much research I had an opinion that artificially intelligent technology is the inevitable future of society so I may as well accept it. After researching the subject I have become more enthusiastic about this upcoming technology because of all the ways it could benefit society. I had a couple of ideas of how AI could be used in electronics or even the classroom but I found a book that told of possible uses in the medical field and industry which I think would be more useful to society as a whole. So to answer my question from earlier about whether the technology is realistic and reliable to perform complex tasks: yes, the technology is in its infant stage right now but will eventually be perfected to where it will be more reliable than humans. Then to answer my other question about whether society should trust computers to think for themselves: I’d respond by saying computers can only do what they are programmed to do and cannot alter their own code. After suppressing the doubts I had in the back of my head about how realistic and reliable artificial intelligence is, and researching the systematic process behind computers’ thought processes to determine if I would trust the technology, I have drawn my conclusion. This technology is in fact inevitably going to be a part of my life in the near future and I think it will benefit me in some way. The benefits this technology would bring are the reasons I am enthusiastic for its arrival and I think everyone should be just as excited.
Work Cited
Ghahramani, Zoubin. “Probabilistic Machine Learning and Artificial Intelligence.” Nature. Nature, 27
May 2015. Web. 25 Feb. 2016.
McEneaney, John E. “Simulation-Based Evaluation of Learning Sequences for Instructional
Technologies.” Springer Link. Instructional Netherlands, 20 Jan. 2016. Web. 25 Feb. 2016.
Winston, P. H., and K. A. Prendergast. The AI Business: Commercial Uses of Artificial Intelligence.
Cambridge: MIT, 1986. Print.