Humanities and Artificial Intelligence: Faculty Seminar
James M. Dolliver NEH
Summer 2023 Faculty Seminar
University of Puget Sound
May 22 - June 2, 2023
Directed by Justin Tiehen (Philosophy) and Ariela Tubert (Philosophy)
Summer 2023 Faculty Seminar
University of Puget Sound
May 22 - June 2, 2023
Directed by Justin Tiehen (Philosophy) and Ariela Tubert (Philosophy)
Details of the seminar:
- 14 Puget Sound faculty members representing 14 humanities and humanistic social science departments and programs
- 25 hours of in-person discussion paired with readings and hands-on activities
Seminar Overview
The word “robot” was first introduced into the modern lexicon not by an engineer, or a mathematician, or a scientist of any sort. Instead, the term was coined by Czech writer and playwright Karel Čapek in his 1920s work R.U.R.: Rossum’s Universal Robots. The play initially presents artificially intelligent beings as humanity’s ultimate achievement. However, after an initial period of serving human beings as perfect workers, the robots in the work rebel, turning against their creators in response to the injustice of being treated as second-class citizens. By the end, with human beings wiped out by the robot rebellion, the play portrays two robots that take on human characteristics of their own, as represented by their willingness to sacrifice themselves for one another. Čapek ultimately uses robots and artificial intelligence as a device for helping the audience better understand their own humanity. This sort of thing is common in the humanities, where authors often imagine possibilities for the future, while at the same time being informed by technology of their day.
This is what our proposed Dolliver seminar is about: the humanities and artificial intelligence (AI). We mean for this to cover both work in the humanities that is about AI, as well as work in the humanities that makes use of or is otherwise informed by AI. It is a glaring fact about life today that AI has infiltrated more and more areas of our lives, as represented by the ubiquity of using Google searches to find needed information, using Amazon’s recommendation algorithm to decide what to purchase, using Facebook or other social media sites, and more. Some of the forms of this AI-infiltration touch especially on matters of humanistic concern–this is the focus of our proposed seminar.
Consider a few examples. Lost Tapes of the 27 Club was an album released in 2021 featuring exclusively music written by AI. The AI system involved was trained on the music made by various artists who all died at the young age of 27–including Kurt Cobain, Jimi Hendrix, Jim Morrison, and Amy Winehousee–and then generated new songs in these artists’ different styles. The idea was to give listeners a sense of what, say, Kurt Cobain might have produced if only he had not tragically died so young, and in the process to raise awareness of the mental health issues that Cobain and other artists faced. Another example is provided by generative art AI systems like DALL-E, Midjourney, and Stable Diffusion, which generate painting-like and photograph-like images in mere seconds after being given words and styles selected by human users. Such projects raise various questions, including whether machines can possess genuine artistic creativity, and just what humanity’s place will be in the creation of arts in the years ahead. Some authors even worry that machines might eventually get so good at making music and painting pictures and writing stories and so on that eventually there will no longer be a place for human artists–audiences will prefer computer-generated art because it will be so much superior. Perhaps there will one day be a robot playwright who puts Čapek and other mere humans to shame, leaving us in a state of post-human humanities.
A different set of issues regarding value and humanistic concern are raised in connection with the burgeoning field of AI ethics. To take one example, the advent of deep neural networks has allowed AI to make extraordinary leaps at identifying people from photographs, as many are aware from the DeepFace algorithm used by Facebook to recognize your friends in the pictures you post. However, the technology raises deep concerns about privacy, algorithmic bias, and more. As a result, in November 2021, Facebook announced plans to shut its program down. Outside of Facebook, some researchers have used facial recognition technology to conduct highly controversial research, with one especially notable example involving a claim to have trained an algorithm to distinguish criminal from non-criminal portraits by attending to certain facial features–given, your lip curvature, say, or your eye inner-corner distance, the algorithm can supposedly predict whether you are more or less likely to commit a crime. Such work has sparked ethical concerns from philosophers and others and has led historians to explore parallels between such AI work today and discredited pseudosciences like physiognomy and phrenology.
Then there is the ongoing work in digital humanities, which makes use of AI to augment rather than replace more traditional humanistic inquiry. For example, recent research by philosopher Mark Alfano attempts to provide an interpretation of Nietzsche’s works by using AI to construct semantic networks and systematically catalog patterns and conceptual schemes in his writing. We cite Alfano because we are especially familiar with his approach, but there are many other examples of similar research being done that uses AI to understand historical texts or materials.
Faculty on campus are already teaching courses that connect the humanities and AI or using AI tools in their humanities research. The Dolliver faculty seminar would draw on their expertise and bring them together with faculty members who are interested but haven’t yet had the opportunity to do so. The goal is to engage members of the seminar into generative discussions about the ethical, social, and existential issues that arise from AI and the ways in which AI can contribute to and enhance work in the humanities. The work in the seminar will connect to the series of public events connected to the topic of AI and the humanities that we will organize in 2023-2024 in collaboration with the Dolliver participants and other members of the campus community. The ultimate goal of the project is to increase the exchange of ideas around issues at the intersection of the humanities and AI on campus, with more visibility in our course offerings and events on campus, so as to ultimately better prepare students to responsibly engage with AI in their lives, work, and research.
The word “robot” was first introduced into the modern lexicon not by an engineer, or a mathematician, or a scientist of any sort. Instead, the term was coined by Czech writer and playwright Karel Čapek in his 1920s work R.U.R.: Rossum’s Universal Robots. The play initially presents artificially intelligent beings as humanity’s ultimate achievement. However, after an initial period of serving human beings as perfect workers, the robots in the work rebel, turning against their creators in response to the injustice of being treated as second-class citizens. By the end, with human beings wiped out by the robot rebellion, the play portrays two robots that take on human characteristics of their own, as represented by their willingness to sacrifice themselves for one another. Čapek ultimately uses robots and artificial intelligence as a device for helping the audience better understand their own humanity. This sort of thing is common in the humanities, where authors often imagine possibilities for the future, while at the same time being informed by technology of their day.
This is what our proposed Dolliver seminar is about: the humanities and artificial intelligence (AI). We mean for this to cover both work in the humanities that is about AI, as well as work in the humanities that makes use of or is otherwise informed by AI. It is a glaring fact about life today that AI has infiltrated more and more areas of our lives, as represented by the ubiquity of using Google searches to find needed information, using Amazon’s recommendation algorithm to decide what to purchase, using Facebook or other social media sites, and more. Some of the forms of this AI-infiltration touch especially on matters of humanistic concern–this is the focus of our proposed seminar.
Consider a few examples. Lost Tapes of the 27 Club was an album released in 2021 featuring exclusively music written by AI. The AI system involved was trained on the music made by various artists who all died at the young age of 27–including Kurt Cobain, Jimi Hendrix, Jim Morrison, and Amy Winehousee–and then generated new songs in these artists’ different styles. The idea was to give listeners a sense of what, say, Kurt Cobain might have produced if only he had not tragically died so young, and in the process to raise awareness of the mental health issues that Cobain and other artists faced. Another example is provided by generative art AI systems like DALL-E, Midjourney, and Stable Diffusion, which generate painting-like and photograph-like images in mere seconds after being given words and styles selected by human users. Such projects raise various questions, including whether machines can possess genuine artistic creativity, and just what humanity’s place will be in the creation of arts in the years ahead. Some authors even worry that machines might eventually get so good at making music and painting pictures and writing stories and so on that eventually there will no longer be a place for human artists–audiences will prefer computer-generated art because it will be so much superior. Perhaps there will one day be a robot playwright who puts Čapek and other mere humans to shame, leaving us in a state of post-human humanities.
A different set of issues regarding value and humanistic concern are raised in connection with the burgeoning field of AI ethics. To take one example, the advent of deep neural networks has allowed AI to make extraordinary leaps at identifying people from photographs, as many are aware from the DeepFace algorithm used by Facebook to recognize your friends in the pictures you post. However, the technology raises deep concerns about privacy, algorithmic bias, and more. As a result, in November 2021, Facebook announced plans to shut its program down. Outside of Facebook, some researchers have used facial recognition technology to conduct highly controversial research, with one especially notable example involving a claim to have trained an algorithm to distinguish criminal from non-criminal portraits by attending to certain facial features–given, your lip curvature, say, or your eye inner-corner distance, the algorithm can supposedly predict whether you are more or less likely to commit a crime. Such work has sparked ethical concerns from philosophers and others and has led historians to explore parallels between such AI work today and discredited pseudosciences like physiognomy and phrenology.
Then there is the ongoing work in digital humanities, which makes use of AI to augment rather than replace more traditional humanistic inquiry. For example, recent research by philosopher Mark Alfano attempts to provide an interpretation of Nietzsche’s works by using AI to construct semantic networks and systematically catalog patterns and conceptual schemes in his writing. We cite Alfano because we are especially familiar with his approach, but there are many other examples of similar research being done that uses AI to understand historical texts or materials.
Faculty on campus are already teaching courses that connect the humanities and AI or using AI tools in their humanities research. The Dolliver faculty seminar would draw on their expertise and bring them together with faculty members who are interested but haven’t yet had the opportunity to do so. The goal is to engage members of the seminar into generative discussions about the ethical, social, and existential issues that arise from AI and the ways in which AI can contribute to and enhance work in the humanities. The work in the seminar will connect to the series of public events connected to the topic of AI and the humanities that we will organize in 2023-2024 in collaboration with the Dolliver participants and other members of the campus community. The ultimate goal of the project is to increase the exchange of ideas around issues at the intersection of the humanities and AI on campus, with more visibility in our course offerings and events on campus, so as to ultimately better prepare students to responsibly engage with AI in their lives, work, and research.