AMT Lab @ CMU

View Original

Using AI Powered Art to Increase Social Equity

Note: This post is Part II of a two part series on the issues surrounding AI and bias, and the role the arts play in addressing these issues. Part I may be found here.

The Spring 2019 Steiner Lectures in Creative Inquiry focused on the intersection between interactive technologies such as AI, VR, and Machine Learning, and the arts. Every year the College of Fine Arts at Carnegie Mellon University brings together professional interdisciplinary artists that use “world-class science and engineering” to talk about their latest research. These professional artists’ research have often led to innovation in the arts industry, which have helped create new policies, redefine artists’ roles, and at times change the art form itself. On Monday, February 4, 2019, I had the privilege of attending Stephanie Dinkins’ lecture, co-presented by the Miller Institute for Contemporary Art, on the uses of Artificial Intelligence to tell stories, increase perception & awareness of each other, and foment social equity within our local communities. During the conversation, Dinkins brought to light some unconventional uses of AI, and posed the audience with a very important question: what if we, as artists, could learn to use AI to remove the barriers between gender, race, and social status? Throughout the talk, she provided insight on utilizing different applications of AI when working with local communities.

The following are the top 3 takeaways for arts managers to consider when thinking of how to better engage audiences and communities within their institutions, as well as some AI applications to consider moving forward:

Dinkins, Stephanie. "NOT THE ONLY ONE, The multigenerational memoir of an African American family told from the perspective of a custom deep learning artificial intelligence." Dinkins Studio. https://www.stephaniedinkins.com/ntoo.html.

AI to tell better stories:

According to Dinkins, Artificial Intelligence can be used to tell better stories which are more personal, relatable, and engaging to audiences attending art institutions. For instance, Dinkins used AI to tell a multi-generational memoir of an African American family in the community through

“the perspective of a custom deep learning artificial intelligence.”

The art piece was called “Not the Only One,” and it captured typical conversations between members of three different generations within the same family. Each side of the sculpture was the face of one of the family members and produced an imitation of their voices talking to each other. To do this, Dinkins interviewed each member of the family and, through AI, collected as much data as possible for the art piece to be able to create its own conversations. The piece was fully interactive, so audience members could walk by and ask the questions they were curious about. Based upon their questions and the AI’s mood, a targeted story from the family would be told. This allowed audience members to be increasingly engaged with the art piece, without the risk of damaging it through touch, while learning about the life and beliefs of an African American family from their home community. The goal, for Dinkins, was to be able to curate real stories from real people within local communities, so that members of the community could start understanding each other at a deeper level. An art piece like this has an enormous potential to increase awareness of the experiences of underrepresented groups of people, and it serves its community as a reminder of their history, and of who the members of their communities are and were.

AI as a Mean for Social Equity and Inclusion:

Dinkins spoke about how technology is still struggling to become inclusive. She pointed at how face recognition technologies are still struggling to recognize people whose skin has darker tones, and how most African American kids in the Brooklyn community had never even tried VR or were aware of AI before she invited them to join one of her recent projects. One of the solutions to this problem was to let the underserved members of each community teach something to developing AI technologies. AI can learn from all of us, however, the reason this technology is not inclusive yet is because of the lack of diversity in the teams who are feeding its networks. In addition, AI technologies tend to be inaccessible to most underserved populations, either because of cost or lack of awareness. Dinkins proposed a solution to these problems through her Project al-Khwarizmi (PAK). Through PAK, Dinkins gathered members from various underserved areas of Brooklyn, NY, a very gentrified city, and invited them to learn about AI technologies and Algorithms, teaching them different ways in which those technologies intersect their day to day lives. The conversation started with an inviting question: What does AI need from you? Once they had an understanding and felt close to the project, the community members got to interact with the AI system in place and feed some of their own culture into its network. Some kids decided to teach AI some hip-hop moves, and others rap lyrics.  While the project is still in progress, Dinkins hopes to make these technologies accessible to all members of the community, teaching them how they work so they can create their own AI technologies and include other members of their communities in the process of AI and machine learning, with the goal of improving AI inclusivity.

Dinkins, Stephanie. "Project al-Khwarizmi (PAK)." Dinkins Studio. https://www.stephaniedinkins.com/project-al-khwarizmi.html.

AI to Expand our Perspective of others:  

As mentioned before, AI can help tell the stories of local community members and help us remember our past, but it can also build perspectives and decrease gentrification and racism. BINA 48 is just one example of how an art piece can help others better understand the perspectives of people who may come from a different background. BINA 48, built by the Tesasem Movement Foundation, is an AI robot modeled after Bina Rothblat, an African-American women, who was designed to capture

“The mind of the real Bina.”

Carnegie Mellon University, School of Art. "STUDIO Lecture: Stephanie Dinkins." http://www.art.cmu.edu/event/studio-lecture-stephanie-dinkins/.

This robot was created with the intention of capturing the essence of a real individual, portraying their thoughts, emotions, and reactions to incoming conversations while removing conversational barriers caused by social discomfort. Dinkins alluded to humans’ natural defense mechanisms, saying that when one feels uncomfortable within a social context, it is common to behave and speak in discordance with one’s true self. As a result, it is difficult for certain community members, especially those who have been a target of discrimination, to express themselves freely. BINA 48 is an example of how the combination of AI and art can be used to remove social constrains and barriers for understanding each other.

From an arts management perspective, BINA 48 has the capacity to learn from conversations had, which offers great potential for arts managers to learn more from their audiences. Designing something similar for arts organizations could potentially multiply the organization’s data collection methods to further their understanding about who is coming to their venues, gain inspiration for designing new exhibits to address the concerns and interests of their audience, and understand what audiences are likely to expect and enjoy when they attend.

The takeaways above are only three examples of how art managers can benefit society starting within their communities, but there are many more out there to find. It will important for managers to keep an eye on Artificial Intelligence and Machine Learning applications in the arts, as they may find it increasingly useful for both data collection and audience engagement.

Worked Cited:

Carnegie Mellon University. "About the STUDIO." http://studioforcreativeinquiry.org/about.

Carnegie Mellon University, School of Art. "STUDIO Lecture: Stephanie Dinkins." http://www.art.cmu.edu/event/studio-lecture-stephanie-dinkins/.

Dinkins, Stephanie. "NOT THE ONLY ONE, The the multigenerational memoir of a black American family told from the perspective of a custom deep learning artificial intelligence." Dinkins Studio. https://www.stephaniedinkins.com/ntoo.html.

Singer, Natasha. "Amazon Is Pushing Facial Technology That a Study Says Could Be Biased." The New York Times, January 24, 2019. https://www.nytimes.com/2019/01/24/technology/amazon-facial-technology-study.html.

Dinkins, Stephanie. "Project al-Khwarizmi (PAK) ." Dinkins Studio. https://www.stephaniedinkins.com/project-al-khwarizmi.html.

STREAMING MUSEUM. "Meet humanoid robot Bina48." http://streamingmuseum.org/bruce-duncan-bina48/ .

Anyoha, Rockwell. "The History of Artificial Intelligence." Science in the News, August 28, 2017. http://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/.