12 GROUNDCOVER NEWS TECHNOLOGY MAY 1, 2026 AI digital transformation: community interests, socio-technological impacts and the future of work From the perspective of ethics and human values, Professor Eric Swanson of the U-M Philosophy Department and Anja Sheppard, PhD student in Robotic AI at the University of Michigan College of Engineering, have shared their ideas. Will: What is Artificial Intelligence (AI)? Anja: Generally, we define artificial intelligence as something that mimics human intelligence, and when people think about AI, they might think about ChatGPT. There is actually a debate on whether this counts as artificial general intelligence because ChatGPT, which is an LLM (Large Language Model), has not accomplished human reasoning. LLMs are very good at generating language. If I ask it to write an essay for me, it would sound very human-like. But the underlying ideas of the essay may not be factually correct or make sense. And this is the reasoning gap that I was talking about. Will: Why do interdisciplinary researchers call AI “digital transformation?” Anja: I think people are thinking about AI as a transformation or a revolution because it is going to substantially change how people live, learn and work. Just think about the internet, which similarly revolutionized the way we live today. Will: What is the difference between Artificial General Intelligence (AGI) and Robotic AI (RAI)? Anja: So, the big challenge for robotics is actually physically engaging with the world. Robots have to safely operate to complete physical tasks. This involves understanding sensory information such as from cameras, microphones, radars and lidars (light detection and ranging). This differs from AI like ChatGPT because robotics have to understand more sensors and they have outputs, with physical commands. ChatGPT takes in text and outputs texts. It is a fundamentally harder problem to design AI for robotics. Why? Because there is far more textbased information on the internet which can help us train AI like ChatGPT than there is data about robots. You cannot use the internet as a source of training data for robot AI. Will: A couple of years ago, ChatGPT made a huge splash in the mass media and the high tech world. Tell us more about what ChatGPT is. What are the benefits? What are the consequences? Anja: ChatGPT is a LLM. It is very good at generating texts. But the drawback is that it is not always trustworthy WILL SHAKESPEARE Groundcover vendor No. 258 because it is prone to hallucination, or making things up, basically. [There are] benefits, for one example, in programming or coding, ChatGPT is very good. This makes programming a lot more efficient, and less tedious. In general, people see it as a tool that improves efficiency. As a consequence, improving efficiency can take away some of the humanity in our work. For example, in the arts, what does it mean to watch a movie written by an AI versus made by human beings? Will: Our readers may want to learn more about how AI would displace many jobs, especially factory work and other occupations such as engineering, writing, journalism, legal researchers, librarians, teachers, movie roles, healthcare or hospitality services. Anja: The jury is still out on this. We are still trying to understand how AI will impact the workforce. There are already reports of entry-level programming jobs being out. So, coming up with strategies to mitigate a huge loss of jobs is really important. The federal government has a job to do: protect people from displacement by regulating the AI industry. Will: What is the future of AI technology in American society within the next two years? Within the next five years? Within the next 10 years? Anja: I do not think I have a good answer for this. I just say that AI will impact pretty much everything we do. Questions asked of Dr. Eric Swanson Will: What are the general and specific impacts on American society of AI? Any hope for Gen Z students who major in computer science? Eric: Great question! Current AI has two impacts on American society, culture, and work that I especially want to highlight. The first is that AI really isn’t the sort of thing that makes commitments to people. Instead, its outputs are more like a prediction that is influenced by Anja Sheppard (left), PhD student in Robotic AI at the University of Michigan College of Engineering, and Professor Eric Swanson (Right) of the U-M Philosophy Department have shared their ideas on AI from the perspective of ethics and human values. what people in general want to hear. By contrast, a person generally has a point of view that involves certain commitments. One person’s point of view involves being a farmer, another person’s involves being a physical therapist, and their expertise and experience influence their commitments. AI isn’t anything like that. It’ll be interesting to see to what extent people embrace or reject a form of life that doesn’t have the kinds of commitments that we do. The second impact involves what the French philosopher Jean-Paul Sartre called ‘seriality.’ When we relate to each other in a ‘serial’ way, Sartre says, we are anonymous to each other and interchangeable with each other. We don’t have any shared project or goals. It’s like waiting for a bus; our actions are organized by waiting in a line, but your project of getting where you need to go isn’t really related to my project of getting where I need to go. We don’t share projects or goals with current AI at all; we treat it serially and its treatment of us has lots of serial properties. Of course, in some ways it’s good for AI to treat us anonymously! But I wonder whether the influence of AI will make people treat each other as interchangeable in new ways. This might make people less inclined to make commitments to each other, and I’m not really sure what that would look like. I myself suspect that it would be a sadder, less meaningful world, but perhaps another interesting perspective is that it would be a freer one. Will: In view of the divergent views of AI from the mass media, what is the public interest? What are the ethical concerns from the perspective of public values? Eric: Another great question, and it interacts with the first! In part because AI doesn’t make commitments and we don’t make commitments to it, and because our interactions with it are relatively serial, it’s pretty easy to just throw it at a problem and see what happens. It’s not like it’ll get mad at you or resent you in any meaningful way! And going back to your first question, you can just tell it that it’s overreacting if it does produce mad sounding text, and close your browser if you want! People do ignore and dismiss each other, but it’s not as easy to do. And people who know you could just ask AI will sometimes resent you not just asking AI. But the drudge work involved with trying and failing to make progress on problems is absolutely crucial to becoming an expert on something. It’s going to be a real challenge to figure out how to balance the availability of a shortcut with the importance of struggling to understand. So much in philosophy is the struggling to understand — better understanding the nuances of the question you started with without necessarily answering every aspect of it. I hope AI use in colleges and universities doesn’t make that seem less valuable than it is. History of AI The online magazine "Tableau" discussed the origin and development of artificial intelligence. It is suggested that AI can be traced to the Ancient Greek period of 400 B.C.: “The idea of ‘artificial intelligence’ goes back thousands of years, to ancient philosophers considering questions of life and death. In ancient times, inventors made things called ‘automatons’ which were mechanical see AI next page
13 Publizr Home