Ethically Safe AI Systems?
Prepared by: Ananta Bhatt- 0878036
Ethical Issues In Computer Science
Artificial Intelligence (AI) field is growing at an unpredictable rate. Even with optimistic approach towards the development, the adverse approach still trails and should not be overlooked. Through this technology, many industries can benefit ranging from the medical field to eliminating wars. AI plays a key role in defining systems that benefit in achieving modular activities. This report highlights the development positive and negative aspects made so far with the technology and the issues in computer ethics governed to decide whether AI systems are ethically safe or not. The pace at which the development in AI is taking place is undeniable and requires attention to developing strong ethically safe AI systems without rendering the natural language processing development. (Cramer, T. ,2018, Winter). It is not far when the robots will actually takeover our world.
With the early growth that initiated in this field involved neural networking concept around 1950s, further stirred excitement for machine learning directives increasing the possibility of creating ‘human-like-thinking’ feature in machines. Moreover, demonstrating the present approach followed in AI comprises of deep learning that lead to boom in this field. With such never ending progress through AI, it becomes a question of matter to decide whether such AI developed are safe today or define malicious use of the technology. Further, ethics cannot be overlooked. Robot ethics comes into the picture while building with moral views an intelligent being.
The quest to develop smart systems that overcome human at all the cognitive tasks sums up to an explosion of intelligence in the systems leaving the human intellect way far behind. The rapid growth in this field, has made it prominent to discuss and understand the good and evil aspects of the evolution in intelligent systems. Further, defining whether these systems are to be considered ethically safe or not.
The ultimate reason for development of AI technology is helping to establish goals that are difficult to achieve by humans such as full-time labour work. Moreover, certain rights such as Robot rights are to be considered for people to be morally obligated towards the machine. However, the underlying factor is the impulsive direction that can led by this technology. The positives and negatives addressed will help provide a clear understanding of its objectives. Analysing these factors will helps us be prepare for the wrong outcomes through this creation. It will help minimise the futuristic catastrophic measures.
One of the positive aspects is Higher end of accuracy. AI systems are trained in order to recognise data and predict the data. This strategy is used in place for processes such as voice recognition, image recognition. From the first program created under AI, the checkers game version that self- learned to play. The predictability of the system to give the desired output helped increase the efficiency and scope of the system. Current innovative system created to predict the lifeline of a person should great aspect of accuracy. About 69% of the predicted cases showed correct mortality rate (Center, healthcare-in-europe, 2016). Another example is of a computer software that defeated the chess champion Garry Kasparov. The computer software is called Deep Junior created by two Israeli programmers Shay Bushinsky and Amir Ban .(Frey, W. P. ,2003).The precision of the computer to respond to the chess move shows the accuracy that the system can achieve. (Bentley, R., 2003, March 20).
These factors are taken into consideration and currently systems are being incorporated to enable AI exploration. Mars 2020 mission will incorporate AI technology to venture the space exploration. Various measures such as digital payments offer personalised experience to consumers leverage the development allowing consumers to perform various processes. For example, cheque photo to transfer or withdraw money. Government is also taking various measures and steps to encourage the use of AI technology such as self-driving cars, robot surgery. Moreover, companies like Google, IBM are creating robust systems encouraging our use of AI. (Dietterich, T. G. 2017, Fall). The development though has many positive aspirations such the system might help us eliminate disease, eradicate wars, and poverty. However, the underlying negative side of the coin still needs to be addressed and look out for. It is prudent to focus now on both the areas before the super intelligent system outsmart us and mark a threat to our civilization.
The fate of our race is entirely dependent on scientists adding new features in order to make our life easier. However, every coin has two sides depending on the side you chose will analyse the growth and progress. There are various organisations and leaders that have come up to spread awareness about the evil fundamentals of this approach. The Future of Life, an Institute also warned that the creation of this technology for areas such as autonomous weaponry. This system will not only help in decreasing the potential number of soldiers that died in the battle but also can cause mass destruction. AI is simple ground-breaking tool developed to benefit our race. However, with the self-learning capabilities that the robots/machines possess makes it inevitable to deny that robots will someday takeover humans. Structurally, the neural networking system is designed as such that it learns from the experience gained and tries to improvise on it. Live example is where the humanlike robot Sophie jokes about taking over the world. (SULLEYMAN, 2017). Can such robots be considered as safe systems?
The core liability lies at understanding the factors that define a system as a safe system. Methods can be designed to continuously monitor the actions of the machines. It will help analyse the cause of the action and help prevent it. Systems that are closely related to our safety must be built with the principles of machine ethics. This will enable the machine system to be solely responsible for their own decisions. (Anderson, M., Anderson, L., S. ,2011). Further, will reduce the probability of negative behaviour. The
Standard ethical morals in machine will help build up the foundation for defining and distinguishing between a right and a wrong decision. Machine ethics functions in a complex dimensional way. But it is important to define the ethical principles for the machines in order for the machines to treat us well enough.
Taking all the concerning factors into account, the positives and the negatives the AI technology is growing at a very fast pace. Even with concerning negative effects that may occur, it is inevitable to slow down its progress. With the development of AI that initiated from 1950s has expanded its reach to all the sectors and is still widening for sectors such as medical, space exploration, mining, searching for fuel. There are subsiding means through which the disastrous effects can be avoided. Ethically defined safe systems can be created in order to avoid the destructive measures that govern machine ethical principles to define logically ethical values.
Dietterich, T. G. (2017, Fall). AAAI Presidential Address: Steps Toward Robust Artificial Intelligence. AI Magazine, 38(3), 3+. Retrieved from Dietterich, T. G. (2017, Fall). AAAI Presidential Address: Steps Toward Robust Artificial Intelligence. AI Magazine, 38(3), 3+.
Frey, W. P. (2003).Chess Skill in Man and Machine. BOOK. Retrieved from https://books.google.ca/books?hl=en&lr=&id=9-3pBwAAQBAJ&oi=fnd&pg=PA1&dq=man+versus+machine&ots=5jyKKJmOR6&sig=8yEVzKFpYPYfaQgsSXcUjzEmLwU#v=onepage&q&f=false
Anderson, M., Anderson, L., S. (2011). Machine Ethics. BOOK. Retrieved from https://books.google.ca/books?hl=en&lr=&id=N4IF2p4w7uwC&oi=fnd&pg=PP1&dq=machine+ethics&ots=5XYTuse_Pk&sig=_mlOJ959iYi1HTGGETR_GSF6uEA#v=onepage&q=machine%20ethics&f=falseBentley, R. (2003, March 20). Man versus machine. Computer Weekly, 24. Retrieved from http://link.galegroup.com.ezpxy.fanshawec.ca/apps/doc/A98970042/PPIS?u=ko_acd_fc&sid=PPIS&xid=ff20388a
Cramer, T. (2018, Winter). The State of Artificial Intelligence. EContent, 41(1), 26+. Retrieved from http://www.econtentmag.com/Articles/Editorial/Feature/The-State-of-Artificial-Intelligence-2018-122577.htm
Artificial Intelligence Transforming the Digital Payments Landscape. (2018, February 21). PC Quest. Retrieved from https://www.pcquest.com/artificial-intelligence-transforming-digital-payments-landscape/
Robitzski, D. (2017, June 2). Inverse. Retrieved from Health: https://www.inverse.com/article/32441-ai-ct-five-years-until-death
SULLEYMAN, A. (2017, July 13). Independent. Retrieved from Indy/Tech: https://www.independent.co.uk/life-style/gadgets-and-tech/news/robots-debate-ai-humanity-take-over-world-lifelike-humans-hans-sophia-a7839061.html
Urban, T. (2015, January 27). WaitButBuy. Retrieved from Ai Revolution: https://waitbutwhy.com/2015/01/artificial-intelligence-revolution-2.html
Center, B. I. (2016, July 11). healthcare-in-europe. Retrieved from Breast Cancer: https://healthcare-in-europe.com/en/story/16749-artificial-intelligence-diagnoses-with-high-accuracy.html