Applications of Artificial Intelligence, Social Robotics
Research report made for my SYDE 261 course: Design, Systems, Society
Introduction
Robots were invented to automate tasks of humans, first meant to be used in the frontline performing tasks in industrial settings. The first autonomous robots were designed to “typically operate in highly structured environments” such as factories and “have only limited interactions with humans” [1]. However now, the next generation of robots - social robots “will have to do something much more complex… fit into everyday human life” [2], and the rise of innovation in artificial intelligence (AI) has made this possible. Social robots is a new development in the field of robotics and an application of artificial intelligence, built to engage with humans in their social environment.
Applications of social robots in the real world include engaging with seniors in old-age homes, childcare, entertainment, cooking, and used in the office as physical counterparts for remote conference participants - referred to as telepresence [3]. Other examples of social robots coming to market include: “bell-hops that deliver extra toiletries to hotel guests, hospitality robots that welcome customers to a store, family assistants that read recipes and keep track of everyone’s calendar, and robotized walkers that enable the elderly to adopt a healthier lifestyle” [1].
Some of the main adoption drivers of social robots include the demand and needs of customers, innovation needs of companies and budgets, simplicity of work processes, average wage for work that these robots are replacing, technological literacy of individuals interacting with social robots, risk to humans (physical and mental) [1]. When organizations have more funds readily available to allocate for innovation are more likely to experiment with social robots in their processes. Simple processes are easily automated with robotics, and there is a higher incentive to replace workers with social robots when those workers have higher wages.
However for these social robots to integrate well in society, “you need a robot that adapts, learns, gets better, responds and tailors its action to a particular interaction partner” [2]. This is no easy feat to design. Not only do designers need to account for technical considerations of building robots, but they will also “need to draw from psychological research in areas such as communication, perception, social-emotional intelligence and theory of mind” and “consider what features robots will need for people to trust them and want to engage with them” [2]. Cynthia Breazeal, an associate professor of media arts and sciences at MIT and leader in social robotics highlighted in a conference presentation that “a social robot should feel much more like you’re interacting with someone rather than a something” [4]. Social robots need to be highly resilient, following an iterative process continuing to learn from their human counterparts and not only have the ability to do tasks but connect with people, recognize emotions, and even make jokes.
Summary of Impacts
The primary users of social robots include the elderly in old-age homes, children (specifically those with autism), and weight loss patients. For the elderly, “robots can engage seniors to foster social connection with other residents in a time in which chronic loneliness is an epidemic with significant health risks” [4]. However, elderly patients may perceive “the use of [a social robot] as an indicator of the loss of some of their previous ability”, potentially creating a stigma and hesitancy to adopt the technology [1]. According to one study, kids who received personalized responses from Tega, a social robot used in the classroom, were more engaged by the activity than those in a control group. In addition, children who related to Tega more had higher vocabulary scores and told longer stories [4]. Specifically for children with autism, “robots allow for a simplified, predictable, and reliable environment, where the complexity of interaction can be controlled and gradually increased” [5]. Caregivers, secondary users of social robots, can easily stay in touch with the elderly through social robots as services can include agenda reminder applications that share information with caregivers so they are aware of the user’s health and safety. For teachers, the individual attention social robots are able to provide saves time for teachers to develop innovative lesson plans that require more cognitive load.
The impact of this can cause the overall literacy in local communities to increase and narrow the education gap of those in marginalized communities if they get equitable access to social robots in educational settings. However, given that social robots are very costly, it is very likely that only the wealthy communities will get first access to social robots, only increasing technological inequality, unless they are funded and mandated by the government.
With the rise of automation, social robots can negatively impact the job market not only for physically demanding roles but now cognitive roles with the power of AI. “Market economics, through the supply and demand curve, suggest that a potential oversupply of people available for these roles will result in lower real wages paid out globally. Workers may be subjected to unpredictable changes in their income” [6]. Acceptance of social robots also vary culturally, with a study measuring negative attitude towards robots scale (NARS), it was found that Japanese participants had a more negative attitude towards their adoption of robots compared to those from the US [7].
The increased use of robots can be evaluated as an environmental burden when considering the materials needed in the material acquisition stage. Since the standardization process of robotic processes of robotic processors and other hardware is limited and usually every new robot is built from scratch, there is a large negative environmental impact during the material acquisition and production phases [8]. Current processes for the end of life of smartphones and other technological devices happen in developed countries by children [9].
As social robots are yet to be widely adopted in homes, for the short-term social robots will take some time to be accepted as a regular part of society. For the long term, Microsoft founder Bill Gates predicted a robot in every home in the near future [7], and the Japanese Robot association predicts that by 2025, the personal robot industry will be worth more than $50 billion a year worldwide [8].
Social Impact and Defensible Recommendation
Social robots’ growing capacity to understand humans and physical appearance becoming similar to humans poses a significant impact on the social robot’s role in human relationships and how companionship is perceived by humans.
With the rise of social robots, there is the risk of dehumanization where humans lose their capacity for love and empathy toward other humans. Robots will give the effect of always giving the user what they want where robots can essentially “be programmed to fulfill every requirement and desire of their user—physically, intellectually and emotionally—making it much more attractive than a human being that has flaws or quirks we cannot simply program away” [10]. Social robots can be programmed to have the same religious beliefs and values as humans, making it difficult to empathize with other human beings who have different religious beliefs than ours. This poses deep ethical concerns to decreasing human’s capacity for emotional depth and potentially moral disgust towards other humans, as they have become too complex in the face of a perfect robotic companion [10].
However, especially due to the COVID-19 pandemic where humans are deprived of human contact, social robots help “fill in gaps to reduce loneliness, depression, and the other serious effects of isolation” [11]. In educational settings, research has shown that there are learning gains “for children being exposed to long‐term interactions with robots that mimic empathic capabilities (such as contingency behaviors and personalization during learning tasks), in contrast with short‐term interactions with robots not endowed with empathic capabilities” [12].
Research has also shown that robots with human‐like appearance can encourage humans to share tasks with a robot, in contrast with machine‐like robots [12]. This thereby proves that there are positive social impacts of having robots with human-like features of empathy, decreasing loneliness, risks of depression, and increasing positive learning outcomes that otherwise would not be achievable with regular mechanical robots.
These positive impacts come with their drawbacks, however. “Several authors have underlined the risk of addiction and attachment to robotic agents (e.g. in the child therapy context) arguing that this can have negative effects relating to regulations and trust” [12]. In addition, since robots do not feel pain but express pain behaviour, this could encourage people to “easily act violently toward sentient creatures, like human beings, who do feel pain while displaying pain behaviour” [10]. Humans, especially children, could be conditioned at a young age to be desensitized to causing pain toward others, changing how we treat other humans.
We must give deep consideration on how we can integrate social robots into society without excluding and replacing emotional connections with humans. Violence towards robots should be unacceptable in society and potentially have government consequences similar to how there are laws on the violence of humans. There should be a limit to the humanization of robots - they should not hold religious beliefs and exhibit facial or sexual features similar to those of romantic partners. Boundaries must be established to uphold human distinctiveness, identity, and value human-to-human interactions for social robots to have a positive social impact.
Environmental Impact and Defensible Recommendations
The increasing adoption of social robots in society poses several threats to the environment. Robots are powered by electricity, and the increased demand for electricity not only to produce the robot but also for these robots to function will greatly increase greenhouse gas emissions, depending on the power source used. Social robots is also a new field with a future of innovation and iteration. It is possible that social robots could potentially fall into the cycle of planned obsolescence, where users are encouraged to purchase newer models of robots every few years because of the rapid change and discovery happening in the artificial intelligence industry. Old robotic parts could potentially fall into the backyards of developing countries or potentially be consumed by endangered species if not discarded correctly. However, on the other hand, social robots have applications that could greatly benefit the environment. Robots allow food to be produced with less labour in the agricultural industry, “maximizing space and production efficiency” [13]. Social robots have the intelligence to monitor pollution output and use data to predict the future of climate change to provide invaluable insight to potentially reduce greenhouse gas emissions [13]. A humanoid robot developed by the Stanford Robotics lab was used to explore the ocean and collect coral reef samples which were essential for a healthy ocean, however, live beyond human reach. The social robot allowed researchers to gain access to samples “without damaging the reef and observe deep-sea specimens in their natural habitat” [14].
Despite the efficiencies and environmental applications of social robots, the “rebound effect” of consumerism suggests that these improvements in cost savings and efficiencies will only create a negative feedback loop from improvements in technology driving an increase in consumption. “With cars, for example, the efficiency savings made by robots have meant more people can afford to buy a new vehicle, increasing the numbers on the roads and the overall emissions from our roads” [13]. Moreover, not all social robots made will have environmental applications as robots made to pick up trash are not considered innovative robots that will drive sales. Encouraging social robots “could also enhance unsustainable exploitation of natural resources—especially if our dependency on rare metals for the production of electronic equipment further deepens” [15]. Companies would rather invest more of their time building social robots that will be implemented in the paying consumer’s homes instead of spending resources on robots for environmental good.
It is evident that social robots can provide both benefits and harm to the environment. To enjoy the innovations of social robots in an eco-friendly matter, it is essential that measures are taken in the production process by using energy-efficient power and recyclable materials to produce the robot. Given that social robots require a lot of material and effort to build, either there should only be a limited number of social robots released to beta test or make old robots easy to upgrade instead of having to completely replace the model with each new iteration. When social robots are at the end of the product life cycle, implementing a recycling program similar to how batteries are recycled can ensure the materials in the robot can be repurposed. Allocating funds by the government to companies can lastly encourage the production of eco-robots.
References
[1] “Social Robots”, KPMG Advisory, Netherlands 2016.
[2] K. Weir, “The dawn of social robots,” Monitor on Psychology, Jan-2018. [Online]. Available: https://www.apa.org/monitor/2018/01/cover-social-robots. An article by the American Psychology Association describing the shift from autonomous to social robots and the scientific challenges of building robots meant to understand people.
[3] J. Myoungh, S. B.Daily, T. Roy, M. T. James, J. Isaac, J. J. Porter, S. S. Darnell, and D. Cherry, “Affective Computing: Historical Foundations, Current Applications, and Future Trends,” in Emotions and Affect in Human Factors and Human-Computer Interaction, Amsterdam: Academic Press, 2017. A paper describing the role of social robots used to inform the introduction of this report. This paper describes the definition of social robots, the use humans have for these devices, and design considerations for social robots.
[4] University of Virginia Darden School of Business, “The Rise of Social Robots: How AI Can Help Us Flourish,” Newswise. [Online]. Available: https://www.newswise.com/articles/the-rise-of-social-robots-how-ai-can-help-us-flourish. An article outlining the positive impacts of artificial intelligence and its applications in social robotics. It talks about its increasing role in our lives and its applications in education and elder care.
[5] C. Breazeal and T. Kanda, “Social Robotics,” in Springer Handbook of Robotics, K. Dautenhahn, Ed. New York City, New York: Springer International Publishing, 2008, pp. 1935–1959. A chapter section with a comprehensive analysis of research done in social robots and how this innovation can help achieve positive outcomes in diverse applications of education, health, quality of life, entertainment and collaboration. It specifically goes into depth of positive impacts for primary users, including children, those with autism, the eldery, and assisted living applications.
[6] P. Golata, “Collaborative robots will soon be in habitual contact with humanity, and their impact will revolutionize the workplace.,” Mouser, 07-Mar-2019. [Online]. Available: https://www.mouser.ca/applications/collaborative-robotics-social-impact/. An article about the social impacts of robots in the workplace, discussing the economics of automation and how jobs could potentially be displaced for older workers with less flexible skill sets.
[7] V. Lim, M. Rooksby, and E. S. Cross, “Social Robots on a Global Stage: Establishing a Role for Culture During Human–Robot Interaction,” International Journal of Social Robotics, 11-Nov-2020. Journal article on the global impacts of social robots exploring the future direction of robots, limitations they currently face, and potential solutions to address these shortcomings. It highlights the critical role culture plays in human-robot interaction.
[8] B. Gates, “A Robot in Every Home”, Scientific American, vol. 296, pp 58-65, 2007. A report written by Bill Gates predicting the future of the robotics industry and how they will be incorporated into society.
[9] A. van Wynsberghe, J. Donhauser, The Dawning of the Ethics of Environmental Robots. Sci Eng Ethics, vol. 24, pp. 1777–1800, 2018. A journal article discussing the environmental impacts of the increased consumption of robots from the material acquisition cycle to the end of use of robotic devices. [10] C. Lanteigne, “Social Robots and Empathy: The Harmful Effects of Always Getting What We Want,” Montreal AI Ethics Institute, 16-Feb-2021. [Online]. Available: https://montrealethics.ai/social-robots-and-empathy-the-harmful-effects-of-always-getting-what-we-want/#:~:text=In%20consequence%2C%20because%20the%20physical,in%20the%20near%20future%2C%20affect. A paper discussing how empathizing with social robots may negatively affect our ability to empathize with other humans, leading to dehumanization. [11] “Pandemic isolation increases acceptance of robot companions,” Waterloo News, 24-Sep-2020. [Online]. Available: https://uwaterloo.ca/news/news/pandemic-isolation-increases-acceptance-robot-companions. An article by the University of Waterloo on how the increasing need for social robot companions is influenced by the pandemic.
[12] J. Giger, N. Piçarra, P. Alves-Oliveira, R. Oliveira, P.Arriaga, Humanization of robots: Is it really such a good idea?, Human Behavior and Emerging Technologies vol. 1, pp. 111-123, 2019. A paper examining the benefits and drawbacks of humanizing robots and the psychological effects of anthropomorphism (making robots increasingly physically resemble and emotionally resemble the characteristics of humans).
[13] L. Wright, “Robots are taking our jobs, but how will this affect climate change?,” Newsweek, 20-Sep-2019. [Online]. Available: https://www.newsweek.com/robots-jobs-affect-climate-change-1460492. An article critically evaluating whether robots really help with climate change with the increased encouragement of consuming technology given the rebound effect. Describes that despite a significant improvement in efficiency, it does not always guarantee an absolute reduction in environmental impact.
[14] A. Dodge, “How Robots Are Helping in the Fight Against Climate Change,” Ozobot, 28-Aug-2018. [Online]. Available: https://ozobot.com/blog/robots-helping-fight-climate-change. An article describing the positive impacts robots have on the environment along with examples of specific robots built for ecological purposes.
[15] J. Dusik and B. Sadler, “What Effect Will Automation Have on the Environment?,” International Institute for Sustainable Development, 22-Jan-2019. [Online]. Available: https://www.iisd.org/articles/automation-environment. Examines the effects automation through artificial intelligence has on the environment providing an analysis on how emerging technologies can lower greenhouse gas emissions along with the importance of understanding how consumption patterns can change through the introduction of emerging technologies.