NEW YORK, N.Y., October 12, 2022 — This week a multinational, virtual audience came together to watch an illuminating discussion featuring several leading powerhouses in the artificial intelligence (AI) ethics space: Dr. Timnit Gebru, founder and executive director of the Distributed Artificial Intelligence Research Institute (DAIR); Minerva Tantoco, Chief AI Officer at the NYU McSilver Institute for Poverty Policy and Research; and moderator Christopher Hemphill, Director of Commercial Intelligence at Woebot Health.
Hosted by The AI Hub at NYU McSilver in its inaugural 2022–2023 Speaker Series event, the conversation delved into how the algorithms that affect our daily lives — from using Siri and Alexa, searching and shopping online, banking and applying for a job; to being surveilled or receiving healthcare services — can place disproportionate harm on Black, brown and other marginalized people, as well as ways they can be used to repair harm. The discussion drew people from Canada, Switzerland, Brazil, Costa Rica, Germany, Australia and the United Kingdom, as well as American cities such as Boulder, CO; San Francisco; Charlotte, NC; Austin and San Antonio, TX; Atlanta; Seattle; Los Angeles; Washington, DC; and the NY–NJ–CT tri-state area.
“Ethical AI is the mindful and intentional design and deployment of AI to serve the public good and reduce its harms,” said Ms. Tantoco. “It is really important to focus not just on the inputs and the data and the output, but also its intentions throughout. What model are you using? How are you testing it? Who’s using it? [And what are you optimizing for?] So, it’s important to have that mindful and intentional approach to the creation of AI and the usage of it.”
Dr. Gebru made headlines in 2020 after Google fired her for raising issues of discrimination in the workplace while she was serving as co-lead of the company’s Ethical AI research team. She said it’s important to question every premise in the development of AI tools, including whether they should even be used at all. Further, Gebru interrogated the expression ‘AI ethics,’ as it also implies “AI for social bad, or AI for social ‘not good.’ Shouldn’t anything be created to be good AI?”
She also cautioned that the pursuit of ethical AI runs counter to the profit motive driving the development of most artificial intelligence. Testing systems to ensure they don’t perpetuate or increase bias can be costlier and take more time. “The institutional incentive structures have to align in order for my work to be impactful… So we either need regulation or something that levels the playing field for everybody.”
Chris Hemphill, who is a 2022–2023 McSilver Fellow-in-Residence, observed, “If we’re asking the same people who are making the systems to self-govern, then there is a major conflict of interest there.”
Ms. Tantoco shared that the way in which The AI Hub at McSilver is putting ethical AI into practice is “to do research using AI techniques to actually point at real problems that no one else is actually looking at, such as issues of racial bias and gender bias. Based on that, we can help to reduce the harms of AI.”
Other topics of discussion included the White House’s AI Bill of Rights, New York’s new regulations on automated employment decision tools and how to encourage more people of color and women to enter the AI development field.
Dr. Gebru is a world-renowned expert in algorithmic bias who is known for taking a stand on the ethical implications of AI. Ms. Tantoco was the inaugural Chief Technology Officer for New York City and holds four US patents on intelligent workflow. Christopher Hemphill hosts the “Hello Health” podcast.
Community Co-Sponsors for the event included the Algorithmic Justice League, Beta NYC, the Distributed Artificial Intelligence Research Institute (DAIR), the New York Hall of Science, the New York Tech Alliance, the NYC Center for Innovation through Data Intelligence, Silicon Harlem, and Women in AI Ethics.
About NYU McSilver
The NYU McSilver Institute for Poverty Policy and Research is committed to creating new knowledge about the root causes of poverty, developing evidence-based interventions to address its consequences, and rapidly translating research findings into action through policy and best practices. It is the home of the AI Hub at McSilver, which investigates how artificial intelligence-driven (AI) systems can be used to equitably address poverty and challenges relating to race and public health. Learn more at mcsilver.nyu.edu.