Artificial Empathy Using Robotics

 

Image of Pepper. Photo Credit: Alex Knight, Pexels

 

By Lilian H. Hill

One example of artificial empathy is Japan's use of robots for elder care. The aging population and a declining birth rate have led to a growing demand for elder care. The national government has invested hundreds of millions of dollars in funding research and development for such care devices using artificial intelligence to display simulations of empathy (Wright, 2023). They are designed to assist in caregiving tasks, provide companionship, and improve the quality of life for the elderly. In addition to robots used for assistive care and safety monitoring, examples of robots endowed with artificial empathy include:

·      Paro: A therapeutic robot designed to look like a baby seal, Paro responds to touch and sound, providing comfort and emotional support to the elderly, particularly those with dementia. The robot is programmed to cry for attention and respond to its name. It includes an off switch.

·      Pepper: Created by Aldebaran Robotics and acquired by SoftBank Robotics in 2015, Pepper is a humanoid robot that can recognize human emotions and engage in basic conversations. It is used in elder care facilities to provide companionship, entertainment, and even lead group activities. Pepper is also used in retail settings for customer service. It talks, gesticulates, and seems determined to make everyone smile.

·      Nao: Originally created by Aldebaran Robotics, acquired by SoftBank Robotics in 2015. Nao is a small humanoid robot designed to interact with people. It is packed with sensors. It can walk, dance, speak, and recognize faces and objects. Now in its sixth generation, it is used in research, education, and healthcare all over the world.

These examples are only a small selection of humanoid robots. For more information, refer to ROBOTS: Your Guide to the World of Robotics (robotsguide.com)

It may strike you as strange, or possibly even creepy, to interact with a robot in intimate ways; however, robots are rapidly being integrated into daily life. The idea of robots was once limited to the world of science fiction, where they were depicted as humanoid machines carrying out tasks with human-like precision and intelligence. Think of R2-D2 and C-3P0 of Stars Fame or Rosey the Robot from the Jetson’s TV Shows. You could also picture Terminator as a more frightening version of movie robotics. Although humanoid robots are still a focus of research and development, robots today come in many different shapes and serve a wide range of functions in our daily lives. Robotics are used in automated vacuum cleaners, Smart home devices, home security systems, and personal assistants like Alexa and Siri (Galiniostech, 2023).

Artificial empathy aims to make interactions with AI systems feel more human-like, fostering trust and comfort in users. However, it also raises ethical considerations about the authenticity of machine-generated empathy and the potential for manipulation.

Wright (2023) notes that there needs to be more connection between promoting robotic care assistants and their actual use. His research in Japan indicates that robotic devices require setup, maintenance, and time to manage and store, reducing caregivers' time with residents. He comments that “existing social and communication-oriented tasks tended to be displaced by new tasks that involved more interaction with the robots than with the residents. Instead of saving time for staff to do more of the human labor of social and emotional care, the robots actually reduced the scope for such work” (para. 13). He concludes by saying the robotic devices may be an expensive distraction from the difficult choices we face regarding how we value people and allocate resources in our societies, leading policymakers to postpone tough decisions in the hope that future technologies will "rescue" society from the challenges of an aging population.

 

References

Galiniostech (2023, November 6). Robots in everyday life: A glimpse into the future. Medium. https://medium.com/@galiniostech/robots-in-everyday-life-a-glimpse-into-the-future-c966640a783d

Wright, J. (2023, January 9). Inside Japan’s long experiment in automating elder care: The country wanted robots to help care for the elderly. What happened? MIT Technology Review. https://www.technologyreview.com/2023/01/09/1065135/japan-automating-eldercare-robots/

 

Artificial Empathy: Creepy or Beneficial?

Photo Credit: Pavel Danilyuk, Pexels

 

By Lilian H. Hill

 

Artificial empathy refers to the simulation of human empathy by artificial intelligence systems, allowing them to recognize, understand, and respond to human emotions in a way that appears empathetic. Empathy encompasses various cognitive and emotional abilities that allow us to understand the internal states of others. Consequently, developing artificial empathy represents both a symbolic goal and a significant challenge for artificial systems, especially robots, as they work towards creating a potentially symbiotic society (Asada, 2018).

Artificial empathy has significant implications for the development of social robots, customer service bots, and other AI applications that interact with humans on a personal level. Below are some key aspects, applications, benefits and drawbacks of artificial empathy.

Key Aspects of Artificial Empathy

Emotion Recognition: AI systems use sensors and algorithms to detect human emotions through facial expressions, voice tones, and body language. These data are processed to identify specific emotional states.

Sentiment Analysis: By analyzing text data from conversations, social media, force and speed of keystrokes, or other sources, AI can gauge the sentiment behind the words and understand the emotional context.

Context Awareness: AI systems are designed to understand the context of interactions, considering factors like the user's environment, past interactions, and specific situations to respond appropriately.

Personalization: Artificial empathy involves tailoring responses based on the user's emotional state and preferences, creating a more personalized interaction.

Behavioral Mimicry: AI can be programmed to exhibit empathy behaviors, such as offering comforting words, showing understanding, or providing appropriate responses in emotional situations.

Applications of Artificial Empathy

Healthcare: AI systems with artificial empathy can support patients by providing emotional comfort, recognizing signs of distress, and improving the overall patient experience.

Customer Service: Chatbots and virtual assistants can use artificial empathy to handle customer inquiries more effectively by responding to the customer's emotional state.

Education: AI tutors can provide personalized support, recognizing when a student is frustrated or confused and adjusting their teaching methods accordingly.

Companionship: Social robots with artificial empathy can provide companionship to individuals, particularly the elderly or those with special needs, by engaging in empathetic interactions.

Benefits and Drawbacks

Artificial empathy can significantly enhance interactions between humans and AI systems but also presents challenges and ethical concerns.

Benefits

AI systems that recognize and respond to emotions create more natural and satisfying interactions, improving user satisfaction and engagement. Empathetic AI in customer service can handle queries more effectively, reducing frustration and increasing loyalty by providing more personalized and considerate responses. AI with artificial empathy can offer support in mental health contexts, providing immediate emotional recognition and support and assisting professionals by monitoring patient well-being. For elderly or isolated individuals, empathetic robots and virtual assistants can provide companionship, reducing feelings of loneliness and improving quality of life.  AI with empathy can be used in educational tools and training programs, providing supportive and encouraging feedback to learners and enhancing their motivation and learning outcomes.

Drawbacks

There is a risk that users may feel deceived if they discover that a machine simulated the empathy they experienced, potentially damaging trust in AI systems.  Emotion recognition often requires sensitive data, such as facial expressions and tone. This raises concerns about data privacy and security and the potential misuse of personal information. AI with artificial empathy could manipulate emotions for commercial or political purposes, exploiting users' emotional states to influence their decisions or behaviors. Over-reliance on empathetic AI for emotional support might reduce human-to-human interactions, potentially impacting social skills and relationships. The development and use of artificial empathy raise ethical questions about the boundaries of human-AI interaction, the role of AI in emotional contexts, and the potential for AI to replace human empathy in critical situations. Current AI systems might misinterpret emotions or provide inappropriate responses, leading to frustration or harm rather than support.

Balancing these benefits and drawbacks is crucial for developing and deploying artificial empathy in AI systems.

 

References

Asada, M. (2018). Artificial empathy. In K. Shigemasu, S. Kuwano, T. Sato, & T. Matsuzawa (Eds.), Diversity in Harmony – Insights from Psychology. Wiley. https://doi.org/10.1002/9781119362081.ch2

Galiniostech (2023, November 6). Robots in everyday life: A glimpse into the future. Medium. https://medium.com/@galiniostech/robots-in-everyday-life-a-glimpse-into-the-future-c966640a783d

Wright, J. (2023, January 9). Inside Japan’s long experiment in automating elder care: The country wanted robots to help care for the elderly. What happened? MIT Technology Review. https://www.technologyreview.com/2023/01/09/1065135/japan-automating-eldercare-robots/

Navigating the Complexities and Dynamics of the Information Ecosystem

 

 

By Lilian H. Hill

 

The information ecosystem refers to the complex network of processes, technologies, individuals, and institutions involved in the creation, distribution, consumption, and regulation of information. It encompasses various elements that interact and influence each other, shaping how information is produced, shared, and used in society. The use of the term ecosystem as a metaphor suggests key properties of environments in which information technology is used. An information ecosystem is a complex system of parts and relationships. It exhibits diversity and experiences continual evolution. Various parts of an ecology coevolve, changing together according to the relationships in the system (Nardi & O’Day, 1999).

 

While the term Information Ecosystem has been in use in academic circles for more than 20 years, it has penetrated today’s media. The dynamic and often unpredictable information ecosystem we inhabit necessitates renewed focus on the fundamental concepts of that ecosystem (Kuehn, 2022). The relationship between information literacy and the information ecosystem is symbiotic and integral. Information literacy refers to the set of skills and knowledge that allows individuals to effectively find, evaluate, use, and communicate information. It encompasses critical thinking and problem-solving abilities in relation to information handling. The term information ecosystem describes the complex environment in which information is produced, distributed, consumed, and preserved. This includes libraries, databases, media, social networks, and other channels and platforms where information flows.

 

Burgeoning and rapidly evolving information technologies influence information production and access. While the emphasis should be on the human activities served by information technologies, the truth is that technology is radically changing ways that information is produced, accessed, understood, and applied.

 

Components of the Information Ecosystem

Multiple constituents work together to produce, distribute, interpret, consume, and regulate information.

 

Information Producers

·      Journalists and Media Organizations: Traditional news outlets, digital news platforms, and independent journalists who gather, verify, and disseminate news.

·      Academic and Research Institutions: Universities, research centers, and scholars who produce scholarly articles, studies, and data.

·      Government Agencies: Institutions that generate reports, statistics, and public records.

·      Businesses and Corporations: Companies that create content for marketing, public relations, and corporate communications.

·      Individuals: Citizens who produce content through blogs, social media, and other personal platforms.

 

Information Distributors

·      Social Media Platforms: Facebook, Twitter, Instagram, LinkedIn, and others that facilitate the rapid spread of information.

·      Search Engines: Google, Bing, and others that organize and provide access to information.

·      Traditional Media: Newspapers, television, radio, and magazines distributing news and entertainment content.

·      Online Platforms: Websites, forums, and blogs that host and share various forms of content.

 

Information Consumers

·      General Public: Individuals who consume news, entertainment, educational content, and other forms of information.

·      Professionals: Individuals in specific fields who create and rely on specialized information.

·      Organizations: Businesses, nonprofits, and governmental bodies that use information for decision-making and strategy.

 

Regulatory Bodies

·      Government Regulators: Agencies that enforce laws and regulations related to media, information privacy, and intellectual property.

·      Industry Groups: Organizations that set standards and guidelines for information dissemination and ethical practices.

 

Dynamics of the Information Ecosystem

Engaging within the information ecosystem requires participating in interrelated activities. Information is generated through research, reporting, personal expression, and other methods. Verification processes, such as fact-checking and peer review, are crucial to ensure accuracy and credibility. Information is distributed through various channels, from traditional media to digital platforms. Access to information is influenced by factors such as digital divide, censorship, and platform algorithms. Individuals consume information based on personal preferences, biases, and social influences. Interpretation of information can vary widely, affecting public opinion and behavior. Consumers provide feedback through comments, shares, likes, and other forms of engagement. This interaction can influence future content production and distribution strategies. Finally, regulatory bodies and ethical standards shape the practices of information producers and distributors. Unfortunately, technological innovations occur more rapidly than regulation and ethical standards. Issues such as misinformation, data privacy, and intellectual property rights are key considerations.

 

Challenges in the Information Ecosystem

With technological advances, numerous challenges exist, including the rapid spread of mis-and dis-information, information overload, echo chambers, inequities, and increased privacy concerns. The spread of false or misleading information can have significant societal impacts, from influencing elections to public health crises. The vast amount of information can overwhelm consumers, making it difficult to discern credible sources.  Algorithms and personalized content can create echo chambers where individuals are exposed only to information that reinforces their existing beliefs. Inequities in access to technology and information resources can exacerbate social and economic disparities. The collection and use of personal data by information platforms raises significant privacy issues.

 

Artificial Intelligence and the Information Ecosystem

AI systems are reshaping the information ecosystem. Information systems play a crucial role in everyday life by influencing and reorganizing people’s thoughts, actions, social interactions, and identities. Hirvonen et al. (2023) argued that the “affordances of AI systems integrated into search engines, social media platforms, streaming services, and media generation, shape such practices in ways that may, paradoxically, result both in the increase and reduction of diversity of and access to information” (p. 1).

 

Fleming (2023) indicated that AI tools can create distorted histories and fake profiles, presenting them persuasively as facts. The stakes are escalating daily as rapid advancements in generative AI pose the risk of escalating online hate speech and misinformation to unprecedented levels. These voices are not new, but the global reach of social media allows lies and conspiracy theories to spread instantly worldwide, affecting millions, undermining trust in science, and fostering hatred potent enough to incite violence. Pernice (2019) indicates that the questions of how to (1) effectively safeguard the deliberative process of building political will and (2) preserve the legitimacy of the democratic process against various IT-driven manipulation attempts remains unresolved. 

 

Importance of a Healthy Information Ecosystem

Peterson-Salahuddin (2023) commented that concerns within information ecosystems include (1) ways information production, particularly in mainstream journalism, can lead to information inequity in its representations and (2) the dissemination and retrieval of this journalistic information via algorithmically mediated online systems, such as social media and search platforms, can replicate and reinforce information inequity within the broader information ecosystem. A healthy information ecosystem is essential for informed citizenship, effective governance, and social cohesion. It promotes:

 

1.    Informed Decision-Making: Accurate and reliable information enables individuals and organizations to make informed decisions.

 

2.    Democratic Participation: Access to diverse and credible information supports democratic processes and civic engagement.

 

3.    Social Trust: A trustworthy information ecosystem fosters social trust and cooperation.

 

4.    Innovation and Progress: Access to knowledge and information drives innovation, education, and cultural development.

 

In a prophetic comment, Nardi and O’Day (1999) indicated that the ecological metaphor conveys a “sense of urgency about the need to take control of our information ecologies, to inject our own values and needs into them so that we are not overwhelmed by some of our technological tools” (p. 49). Maintaining a healthy information ecosystem requires efforts from all stakeholders, including information producers, distributors, consumers, and regulators, to uphold standards of accuracy, fairness, and transparency.

 

References

Fleming, M. (2023, June 13). Healing Our Troubled Information Ecosystem. Medium. https://melissa-fleming.medium.com/healing-our-troubled-information-ecosystem-cf2e9e8a4bed

Hirvonen, N., Jylhä, V., Lao, Y., & Larsson, S. (2023). Artificial intelligence in the information ecosystem: Affordances for everyday information seeking. Journal of the Association of Information Science Technology, 74(12), 1–14.

Kuehn, E. F. (2022). The information ecosystem concept in information literacy: A theoretical approach and definition. Journal of the Association of Information Science Technology, 74(4), 434-443. https://doi.org/10.1002/asi.24733

Nardi, B. A., & O’Day, V. L. (1999). Information ecologies: Using technology with heart. MIT Press.

Pernice, I. (2019, March 5). Protecting the global digital information ecosystem:  A practical initiative. Internet Policy Review. https://policyreview.info/articles/news/protecting-global-digital-information-ecosystem-practical-initiative/1386

Peterson-Salahuddin, C. (2024). From information access to production: New perspectives on addressing information inequity in our digital information ecosystem. Journal of the Association for Information Science & Technology, 1. https://doi-org /10.1002/asi.24879 

 

How Artificial Intelligence Influences Voters and Election Results

 

Image Credit: Edmond Dantès, Pexels

Artificial Intelligence (AI) tools can directly influence voters through the widespread adoption of chatbots integrated into search engines. In this podcast episode, Dr. Lilian Hill discusses how AI influences voters and election results. 

 

References

Noti, A. (2024. February 28). How Artificial Intelligence Influences Elections, and What We Can Do About It. Campaign Legal Center. https://campaignlegal.org/update/how-artificial-intelligence-influences-elections-and-what-we-can-do-about-it

Panditharatne, M. & Giansiracusa, N. (2023, Juy 21). How AI Puts Elections at Risk — And the Needed Safeguards. Brennan Center for Justice. https://www.brennancenter.org/our-work/analysis-opinion/how-ai-puts-elections-risk-and-needed-

 

 

Listen to the Podcast

 

Information Literacy Episode 23 Transcripts

 

 

Making Sense of Complexity: Typologies of Artificial Intelligence

 

Image Credit: Microsoft Stock Images

    

Artificial intelligence (AI) influences many aspects of modern life and has multiple applications. AI is the ability of machines or software to perform tasks that are commonly associated with human intelligence, such as recognizing patterns, making decisions, or learning from data. AI is designed to mimic human capabilities, including pattern recognition, data analysis, and decision-making, and to perform tasks rapidly and efficiently.

 

Algorithms are a set of problem-solving steps computer programs use to accomplish tasks. AI operationalizes the algorithmic steps in smart machines that perform tasks usually associated with human intelligence such as “learning, adapting, synthesizing, self-correction, and use of data for complex processing” (Popenici & Kerr, 2017, para. 3). Machine learning is an application of AI in which large data sets are analyzed, without direct instruction, to detect patterns that might elude human beings. Generative AI is an artificial intelligence technology that can produce various types of content, including text, imagery, audio, and synthetic data.

 

AI originated in the 20^th century, but only recently have computers had the computational power to make it practical and useful (Anyoha, 2017). Most people are using AI without recognition because AI powers internet search platforms, predictive text, grammar- and spell-check, GPS, social media curation, smart devices, streaming services, and patient portals. Many people conflate generative AI with large-language models such as those used within ChatGPT, but this is only one type of AI.

 

Typologies of AI

It is essential to recognize that AI is multifaceted and has multiple applications. Therefore, it can be categorized in multiple ways: based on capability, functionality, application, or degree of supervision vs. autonomy.

 

Capability

One capability-based categorization is weak and strong or general AI. Narrow or weak AI can perform single-specific tasks such as making Netflix recommendations, facial recognition, self-driving cars, searching the internet, or translating languages. General or Strong AI can perform tasks in a human-like manner (AVContent Team, 2023). Some descriptions differentiate general AI from strong AI, with the former referring to a computer that is as smart as a human in a general sense and the latter referring to computers that have achieved human consciousness. The latter category is still somewhat theoretical because AI has not yet achieved consciousness or self-awareness. This counteracts the idea that sentient robots will take over the world and enslave humans, as many science fiction novels and films would have people believe. Think of the Terminator or 2001: A Space Odyssey. 

 

Another capability-based typology characterizes four levels of AI: (a) reactive, (b) limited memory, (c) theory of mind, and (d) self-awareness. Consistent with the weak vs strong typology, this conceptualization indicates that AI has not yet achieved theory of mind, meaning the capacity “to understand and remember other entities' emotions and needs and adjust their behavior based on these. This capability is like humans in social interaction” (Arya, 2023, para. 13). Humans develop this capacity as they mature. They also develop self-awareness and emotional intelligence, while AI does not.

Four Levels of AI

Functionality

One functionality categorization scheme asserts three categories of AI: (a) large language models (LLM), (b) learning analytics in which personalized learning is tailored for individuals, and (c) big data, meaning using large data sets to conduct comparative analysis between groups of people. These can be expressed in input and output, instructor and student, or data and functions.

Another functionality schematic suggests the following categories:

·       Analytic AI scans large datasets to identify, interpret, and communicate meaningful patterns of data.

·       Functional AI scans huge amounts of data to take actions.

·       Interactive AI automates communication without compromising on interactivity.

·       Text AI uses semantic search and natural language processing to build semantic maps and recognize synonyms to understand the context of user’s question.

·       Visual AI identifies, recognizes, classifies, and sorts objects or converts images and videos into insights. (Sarker et al., 2022).

Application

Yet another way of categorizing AI is by applications in which it is used. For example, expert systems use information collected from recognized domain experts to facilitate fast decision-making. Natural language processing (NLP) enables AI to use language in a human-like manner in chatbots, language translation, and sentiment analysis, which is used to determine whether the emotional tone of a message is positive, negative, or neutral. Sentiment analysis has become an important business function used to improve customer service, market research, and to monitor brand performance. It can distinguish the positive from the negative of a seemingly contradictory sentence such as: “While I liked this product, I was disappointed with the color.”

Supervision vs. Autonomy

This typology is often used to describe the process of machine learning, in which:

·       Supervised learning—all data are labeled.

·       Semi-supervised—some input data are labeled, while some are not.

·       Unsupervised—all input data are unlabeled (Alloghani et al., 2020).

 

These terms can also be used to describe AI. Examples of supervised processing include virtual assistants such as Siri and Alexa, while an example of unsupervised or autonomous processing is self-driving cars.

You may notice overlaps between the different typologies, as the following concept map clarifies.

No matter how you conceptualize it, the field of AI is complex, growing, and rapidly being integrated into multiple fields of professional practice. These typologies highlight the diverse nature of AI, and the various systems designed for specific purposes and possessing different levels of capabilities. The field of AI continues to advance and new typologies may be developed as its capacities evolve.

 

References

Alloghani, M., Al-Jumeily, D., Mustafina, J., Hussain, A., Aljaaf, A.J. (2020). A systematic review on supervised and unsupervised machine learning algorithms for data science. In M. Berry, A. Mohamed, & B. Yap (Eds.), Supervised and unsupervised learning for data science. Unsupervised and Semi-Supervised Learning. Springer, Cham. https://doi.org/10.1007/978-3-030-22475-2_1

 Anyoha, R. (2017, August 28). The history of artificial intelligence. Harvard University. Retrieved https://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/

 Arya, N. (2023, November 16). Theory of mind AI in artificial intelligence. Ejable. Retrieved from https://www.ejable.com/tech-corner/ai-machine-learning-and-deep-learning/theory-of-mind-ai-in-artificial-intelligence/#:~:text=Theory%20of%20Mind%3A%20This%20is,like%20humans%20in%20social%20interaction.

AVContent Team (2023, September 14). Weak AI vs strong AI: Exploring key differences and future potential of AI. Analytics Vidhya. Retrieved https://www.analyticsvidhya.com/blog/2023/04/weak-ai-vs-strong-ai/

 Popenici, S. A. D., & Kerr, S. (2017). Exploring the impact of artificial intelligence on teaching and learning in higher education. Research and Practice in Technology Enhanced Learning,12, 22. https://doi.org/10.1186/s41039-017-0062-8

 Sarker, I.H.  (2022). AI-based modeling: Techniques, applications and research issues towards automation, intelligent and smart systems. SN Computer Science,. 3, 158. https://doi.org/10.1007/s42979-022-01043-x