Running Head: AI and Robotics 1. The Applications of AI: The Uses of AI in the Field of Robotics. Michael Lam IT Professor Richard Reo

Transcription

1 Running Head: AI and Robotics 1 The Applications of AI: The Uses of AI in the Field of Robotics Michael Lam IT Professor Richard Reo 17 February 2017 Honor Statement: "By placing this statement on my webpage, I certify that I have read and understand the GMU Honor Code on and as stated, I as student member of the George Mason University community pledge not to cheat, plagiarize, steal, or lie in matters related to academic work. In addition, I have received permission from the copyright holder for any copyrighted material that is displayed on my site. This includes quoting extensive amounts of text, any material copied directly from a web page and graphics/pictures that are copyrighted. This project or subject material has not been used in another class by me or any other student. Finally, I certify that this site is not for commercial purposes, which is a violation of the George Mason Responsible Use of Computing (RUC) Policy posted on web site." The GMU Honor Code statement above will be left justified and double spaced. Hyperlink the GMU Honor Code and George Mason Responsible Use of Computing (RUC) Policy links.

2 AI and Robotics 2 The Applications of AI: The Uses of AI in the field of Robotics Introduction Do you know what runs the smart cars on the road or the Siri/S-Voice that is on your smartphone? The answer: AI or Artificial Intelligence. Since the creations of AI and as it developed, Artificial Intelligence has become a widely-used concept and applied to many different technologies on either small or large scales. One of the main uses for AI is to operate Robotics. AI is already being applied to a lot of things like the Roomba that cleans your rooms or the cars that have sensory notifications when cars are in your blind spot. However, alongside applications of AI, there are also some security, ethical/legal, and social issues that might arise. In the future, once the applications of AI in Robotics improve, it can have a major impact in modern society like robots in the workforce or robots in the military. Background/Current Use The term, Artificial Intelligence, was coined by John McCarthy of Stanford University in In 1958, his paper that was published, Programs with Common Sense, was regarded by many as the first on logical AI (Bogue, 2014). AI is a sub-unit of computer science. The goal of AI is to enable the development of computers that are able to do things normally done by people. There are multiple kinds of AI: Strong AI, Weak AI, General AI, and Narrow AI. trong AI is "aimed at genuinely stimulating human reason in that any result can be us3d to not only build systems that think but also to explain how humans think as well" (Hammond, Kris) whereas, Weak AI is "aimed at just getting systems to work" (Hammond, Kris). General AI is AI that is able to reason and react to external influences so that it can respond accordingly. Narrow AI is AI that is only able to perform it's function only regardless of what external influences happen or

3 AI and Robotics 3 change. Within the subject of AI, it goes further in depth with Machine Learning and Cognitive Learning. Machine Learning involves the design and construction of computer applications or systems that can learn based on their data inputs and/or outputs. Cognitive Learning is where after it gathers information to learn using the Machine Learning, it takes the information of data, patterns, or situations, or the combinations of the three, to reason creatively to make decisions on things that have not happened yet (Machine Learning and Cognitive Systems). Currently, AI is used for subjects like: Virtual Reality, Game AIs, and Robotics and are useful in different fields. In the medical field, Virtual Reality (which is under the concept of AI) is used as a tool for VRET (Virtual Reality Exposure Therapy) for PTSD patients (Senson, 2016). Therapists use the virtual reality to allow PTSD patients to be exposed to their traumatic experience in a safe and controlled environment. AI is also used in many videogames today. Creators of video games uses AI to move NPCs (Non-Playable Characters) or even to get them to react to certain situations in the game. Alongside these uses for AI there are two fields of Robotics that embody the greatest number of AI concepts: humanoid robots and autonomous, mobile robots. For humanoid robots, the best known is Honda s ASIMO (Advance Step in Innovative Mobility) (Inside Asimo). It has the ability to recognize moving objects, postures, gestures, its environment, sounds, and faces which allows it to interact in a limited manner with humans. For autonomous robots, auto and semi-autonomous robots and robotic vehicles are being developed to conduct very specific tasks such as search and rescue operations and a range of military applications (Bogue, 2014). In Japan, they recently opened a robot hotel where almost the entire staff are robots. In the Henn na Hotel, which translates to strange hotel, there are robots working the check-in desk, robotic bag-check, and even a robot concierge (Japan s Robot Hotel).

4 AI and Robotics 4 Security Aspects Some security concerns that the concept of AI can have is that since it is based on technology, it is hackable by anybody who can get past its programming. Also, since AI is technology-based, it is prone to EMPs which can shut it down completely. Another security concern that robots can have is that most of the robots used today are single-task programmed, meaning that the robots are only meant to do a certain job. So, because on this programming, if anything goes off the normal, the robot would not know what to do or will continue to do their job even if it is considered wrong. A real-life example of this is the death of a plant worker in Germany who was crushed by a robot arm designed to install car parts (Frankenstein s Children). Although the robots may be sophisticated enough to do their job, they can make poor decisions because they lack the human ability to think outside of the box or react to unknown variables. Ethical/Legal Aspects Some ethical problems/legal aspects that AI could have is that it could be used against the user if malfunctions occur (Frankenstein s Children). For example, unless the coding for robots are 100 percent accurate and run smoothly, there could be malfunctions that makes the robot attack citizens. Since sophisticated robots are a relatively new concept, they do pose issues to law, regulations, and ethics. Because of this, a study group was created at Harvard University to examine the Law and Regulation of Emerging Robotics and Automation Technologies. In their study group they examined robot and automation technologies that are to be released to the mass public such as self-driving cars or robot assistants. The study group had four sessions with their own topics that raises legal/ethical questions about robotics and autonomous technologies: 1. What is a robot or robotic machine, for purposes of law, regulation, and ethics?

5 AI and Robotics 5 2. Design issues in relation to human cognitive response: a. How should social robots systems be designed and engineered, and how should they be conceived in various social settings. b. How should the problems of incomplete automation be addressed? 3. Design issues in relation to human affective response: a. Trust and reliance issues potentially created by robot design is it possible that different types of machine design might result in a robot hat under-elicits or overelicits human trust and reliance with respect to some particular function? b. Are there legal, regulatory, ethical, or therapeutic issues raised by social robots that elicit in some people a tendency to withdraw from ordinary social interactions with people. 4. Accountability, liability, responsibility, tort actions, and insurance the legal, regulatory, and ethical issues of human accountability for what machines do, or fail to do. How is accountability to be maintained through law and regulation but without tossing out the benefits of the new technologies that will, like all technologies, result in some level of accidents and failures (Law and Regulation). Unfortunately, the study group is relatively new so the results of their sessions have yet to be released. However, examining the topics alone can produce some ideas as to what direction their conversation might be heading to. For the first topic, the clear legal issues that robots and autonomous technologies face is, in the eyes of the law: What exactly is, or could be classified as, a robot or robotic machine? To some, a robot could be a well-sophisticated machine that can interact with society with accuracy. To others, a robot could be a toy dog that runs on small batteries that barks whenever someone pets it s head. So the need to distinguish/clarify what

6 AI and Robotics 6 would be considered a robot in the law is a necessity. The second topic is focuses on the design of the robot/autonomous technology. Should the robot be designed so that it appeals to the public or should it be designed to fit its function in society? Also, topic two questions the ability to detect faults in automatic operating systems, like in an airplane with autopilot that suddenly does not work. The third topic also focuses on the design but in relations to how a person might behave towards it. It questions whether the design might affect the trustfulness of the robot and if, by having a robot, it desensitizes the user to interacting with robots more and with other humans less. The last topic focuses on accountability. Who would be held accountable if a robot malfunctions. For example, if a person in a self-driving car uses auto-pilot but then gets into a car accident, who would be held accountable? The man who was in the car because he relied too much on the self-driving capabilities, or the car manufacturer or AI developer of the car that gave it the auto-pilot function for not making the auto-pilot completely safe? Social Implications Some social implications that AI technology could have is when AI is applied to robots. Once AI technology becomes more advanced in the robot industry, there could be a possibility for the robots to completely take over jobs from the public. An example of this is with the Robot hotel in japan. A reporter asked, Hideo Sawada, the man in charge of the hotel, Isn t hospitality about connecting with people isn t that an important part of the hotel business. (Japan s Robot Hotel) In Sawada s business point of view, he answered For five-star hotels that are selling high-end service, human staff are essential But for three or four star hotels, you need comfortable lodging, and a basic level of communication at a reasonable price. (Japan s Robot Hotel) By having robot as a staff, Sawada says that. in the long term, the payroll for robots (repair/improvements of the robots) are far less than having a payroll for human

7 AI and Robotics 7 employment (salaries/wages). When it comes to having robots in the workforce, especially in fields of hospitality and nursing, the main concern is the connection aspect with the customer/patient. In nursing, the maids are generally the ones who interact with the patients the most, either by assisting the patient in any way or simply having a conversation with a patient. However, once/if robots are brought into the nursing field, that human connection between patient and nurse disappears. The robots can assist with helping the patient in any way possible but robots, as of yet, cannot have a full conversation with the patient. The same can be applied to hotel hospitality. And, if robots become part of societies everyday lives, we will rely too heavily on robots that we might not be able to do anything on our own. Future Uses Some future uses for AI technology are: Robots in the work force, Robots in warfare, robots in entertainment, and etc. Like the concepts of robotics in the movie, Wall-E, once AI becomes more sophisticated, we might be able to have robots assisting humans in our everyday lives and perform jobs that many humans would not want to do like garbage disposal. In terms of warfare, there currently exists technology where a person can control the movements of robots with their mind. If this technology improves and becomes more easily accessible, it could eventually be used in warfare where instead of having human soldiers fight in wars, the robots would be the ones in the field which are controlled by soldiers back at a military base. With the hologram fad that been prevalent in Japan in the entertainment industry, a future perspective that japan plans on doing is to create a strong AI that, when applied to robots, can make the robot sing and dance (Nishida, 2012). To be able to do this, it would require better AI coding because it would require technical issues such as synthesizing natural singing voices and/or body/facia

8 AI and Robotics 8 movements [and] artistic issues such as producing emotional and aesthetic representations to entertain people (Nishida, 2012). Conclusion The age of AI technology has just begun and will continue to improve as other technologies advance. Once it improves, it opens the possibilities of being used in our everyday lives. Even now, AI has been integrated into our lives. 3D printers exist to create objects out of nothing by simply inputting a design for the AI to recognize and make. Finger-print recognition is now a common feature that smart-phones have. Auto-pilot on airplanes for the pilots to use on those long flights or even self-driving cars like Tesla or the Google car. Granted, there are still AI technologies like robotics that are not the best yet. Among the robot technologies that exist today, there are some (a) security issues, (b) ethical/legal issues, and (c) social issues that are always possible. One day, once modern technology, AI, and robotics become more developed, it is possible that in the future, there will be new innovations that can save lives as well as make jobs more efficient.

9 AI and Robotics 9 References Artificial Intelligence, Real Security Problems? Meet Frankenstein's Children. (2016, January 20). Retrieved February 12, 2017, from In this source, it talks about the the possibilities of AI robots shaking off the yoke of their human masters and rebel. It explores the idea that as humans keep making AI smarter and smarter, it will eventually become smart enough that it no longer needs to listen to or even needs humans to sustain itself, naming it to the Frankenstein s Monster scenario. The source gives an example of how realistic AI is becoming in robots by referring to Singapore. Researchers in Singapore have developed an AI receptionist that looks fully human a first glance. It can perform social interactions such as shaking hands, making eye contact, and responding to simple queries. Throughout this source it gives other examples of AI in robotics that deem to be too realistic for comfort. This source is relevant to my research paper because it gives insight on some security, social, and legal issues that might come from the further development of AI. Bogue, R. (2014). The role of artificial intelligence in robotics. The Industrial Robot, 41(2), doi:

10 AI and Robotics 10 In this source, it initially introduces the concept of AI, explaining where the term came from, and how the idea came to be. It gives credit to Jon McCarthy of Stanford University in 1956 for being the creator of the idea of AI and that his rationale was that a central property of humans, intelligence, can be precisely described in code to be simulated by a machine. This is important and relevant to the use of this paper because it gives background of the concept of AI as well as how it is applied in the area of robots. The applications of AI fall into two fields of robots: humanoid robots and autonomous, mobile robots. Inside Asimo. (n.d.). Retrieved February 27, 2017, from This source is an interaction website on the Honda ASIMO, the world s most advanced humanoid robot. Depending on what is clicked, the viewer can view information on the ASIMO like what it can do, how it is made, and Honda s previous innovations. Even when entering the site, the voice of the ASIMO is heard giving narration about what the viewer can do on the site. This source is important because it a clear example of the application of AI. It demonstrates just how sophisticated AI in Robots can get and what it takes to get to that level of development. Law and Regulation of Emerging Robotics and Automation Technologies: Study Group. (n.d.). Retrieved February 12, 2017, from This source is a information page for a study group in The Berkman Center for Internet & Society, exploring the law and regulations of emerging robotics and automation technologies. The source gives bullet points that need to be addressed when it comes to robotics and AI since it is an ever-changing technology. It is split into 4 topics: The first asks, What is a robot or robotic machine, or purposes of law, regulation, and ethics. The second topic questions the design of

11 AI and Robotics 11 the robot in relation to human cognitive response. The third topic questions the design in relation to human affective responses. And the last topic talks of accountability, liability, responsibility, and insurance for the actions of said robot. Who would take responsibility for the actions of the robot and such cases like that. This source is useful to my research paper because it clarifies of some possibly legal issues that might come from the development of robots and AI. Machine Learning and Cognitive Systems: The Next Evolution of Enterprise Intelligence (Part I). (n.d.). Retrieved February 12, 2017, from This source goes further in depth of the components of AI. It goes from the general concept of AI to the concepts that are a deeper understanding of AI and it components like Machine learning and Cognitive systems. In it, it explains what each component is. Machine Learning: involves the design and construction of computer applications or systems that can learn based on their data inputs and/or outputs. Cognitive Learning, is the case where after it gathers information to learn using the Machine Learning, it takes the information of data, patterns, or situations, or the combinations of them, to reason creatively to make decisions on things that have not happened yet. This source is useful to my research paper because it give background knowledge on how AI works and it s general components so that others can understand exactly what AI is and what it takes to develop AI. News, C. (2015, July 22). A night in Japan's robot hotel. Retrieved February 18, 2017, from In this news source, it examines Japan s recently opened robot hotel and gives some details about it from a reporter who actually got to experience staying at the hotel. The reporter

12 AI and Robotics 12 briefly interviews Hideo Sawada, the man in charge of the hotel, about the reasons why he has robots in the hotel and briefly on how they work. This source is useful because it gives a clear example of a new use of robotics in the workforce. Also, it sets an example of how robotics might be used in the future once AI improves. Finally, with this news source, it helps raise a few questions about the social aspect of having robots in the work force. Nishida, T. (2012). The best of AI in japan - prologue. AI Magazine, 33(2), Retrieved from In this source, it is the first report of the best AI accomplishments in Japan. It first gives background knowledge on how AI began to develop in Japan starting from when it first became popular in the 1970 s. From then on, it began to grow from the first Katakana to Kanji conversion system to creating an organization called the JSAI (The Japanese Society for Artificial Intelligence). A future perspective that this source gives that japan plans on doing is to create a strong AI that, when applied to robots, can make the robot sing and dance. This is relevant to my research paper because it explains how complex AI technology can get, especially when applied to robotics because to make a robot dance and sing like a human, it requires technical issues such as synthesizing natural singing voices and/or body/facia movements [and] artistic issues such as producing emotional and aesthetic representations to entertain people. Senson, A. (2016, January 06). Virtual Reality Therapy: Treating The Global Mental Health Crisis. Retrieved February 27, 2017, from

13 AI and Robotics 13 This is source explains the uses of Virtual Reality in the Medical Field. This source focuses on the use of VR in post-traumatic stress therapy. In the source, it talks about using the VR to allow PTSD patients to experience/overcome their traumatic experience in a safe and controlled environment until it is not a stressful event anymore. This source is important to my research paper because it is an example of a current use of AI which is relevant to my paper. It focuses on the psychological aspect of using VR for the therapy of PTSD patients such as visual stimulation and, when combined with headphones, auditory stimulation. Hammond, Kris. "What Is Artificial Intelligence?" Computerworld. Computerworld, 10 Apr Web. 06 Apr