To give female academics and others focused on AI their well-deserved and overdue spotlight time, TechCrunch is launching a series of interviews highlighting notable women who have contributed to the AI revolution. Start. As the AI boom continues, we'll be publishing several articles throughout the year highlighting key research that may go unrecognized. Click here for a detailed profile.
Rashida Richardson is a senior advisor at Mastercard, where she is responsible for legal issues related to AI as well as privacy and data protection.
Mr. Richardson is director of policy research at the AI Now Institute, a research organization that studies the societal impacts of AI, and has previously served as senior policy advisor for data and democracy in the White House Office of Science and Technology Policy. He is currently an assistant professor of law. He will study political science at Northeastern University starting in 2021, where he specializes in race and emerging technologies.
Rashida Richardson, Mastercard AI Senior Advisor
In short, how did you get started in AI? What attracted you to this field?
My background is as a civil rights attorney, working on a variety of issues including privacy, surveillance, school desegregation, fair housing, and criminal justice reform. While working on these issues, I witnessed the early stages of government adoption and experimentation with AI-based technologies. In some cases, I helped lead a number of technology policy efforts in New York State and the City to establish stronger monitoring, evaluation, or other safeguards because the risks or concerns were clear. Did. In other cases, I have focused on making the case for the benefits and effectiveness of AI-related solutions, particularly those marketed to solve or alleviate structural problems such as school desegregation and fair housing. I was skeptical.
Furthermore, from my past experiences, I have become acutely aware of the gaps between existing policies and regulations. I quickly realized that there were very few people in the AI field with my background and experience, or who could provide the analyzes and potential interventions that I was developing in my policy advocacy and academic work. . I then realized that this was a field and space where I could make a meaningful contribution and build on my previous experience in a unique way.
I decided to focus both my legal practice and academic research on AI, specifically the policy and legal issues surrounding its development and use.
What work (in the AI field) are you most proud of?
We are pleased that this issue is finally receiving more attention from all stakeholders, especially policy makers. In the United States, there is a long history of laws either trailing or failing to adequately address technology policy issues, and five or six years ago it felt like that might be the fate of AI. Rather than taking place in formal forums such as U.S. Senate hearings or education forums, many policymakers view the issue as esoteric or urgent, despite the rapid adoption of AI in various fields. I treated it as something I didn't need. However, over the past year or so we have seen a major shift, with AI becoming a constant feature of public debate and policymakers becoming more aware of the risks and the need for informed action. It was done. I also think stakeholders across all sectors, including industry, recognize that AI has unique benefits and risks that may not be resolved through traditional practices. So I think there's a growing awareness, or at least appreciation, of policy interventions.
How do we overcome the challenges of a male-dominated tech industry and, by extension, a male-dominated AI industry?
As a Black woman, I am used to being a minority in many fields, and the AI and technology industry is not new, although it is a very homogeneous field, with huge power and wealth such as finance and finance. It's not that different from other fields. Legal profession. So my previous work and lived experience has prepared me for this industry, as I am fully aware of the preconceptions that I may have to overcome and the difficult dynamics that I will likely encounter. I hope it was helpful. I have a unique background and perspective working on AI across all industries, including academia, industry, government, and civil society, so I rely on my experience to guide us.
What issues should AI users be aware of?
Two important issues that AI users should be aware of are (1) a better understanding of the capabilities and limitations of various AI applications and models, and (2) how current and future legislation can resolve disputes and specific concerns. There is great uncertainty as to whether or not. About AI utilization.
On the first point, there is an imbalance in public discussion and understanding of the benefits and potential of AI applications, as well as their actual capabilities and limitations. This problem is further exacerbated by the fact that an AI user may not understand the difference between an AI application and a model. The release of ChatGPT and other commercially available generative AI systems has increased public awareness of AI, but these AI models are in line with other types of AI models that consumers have been using for years, such as recommendation systems. is different. When conversations around AI are muddy (when the technology is treated as a monolith), public understanding of what each type of application or model can actually do and the risks associated with its limitations and shortcomings is distorted. There is a tendency to
Second, laws and policies regarding the development and use of AI are evolving. Various laws already apply to the use of AI (civil rights, consumer protection, competition, fair lending, etc.), but we are not sure how these laws will be enforced and interpreted. It's still in its early stages. We are also in the early stages of AI-specific policy development. However, what I have found both in my legal practice and in my research is that this legal patchwork leaves unresolved areas that will only be resolved with further legislation. Litigation regarding the development and use of AI. Legal uncertainties about the current state of the law and AI in general, and important issues such as liability, may result in certain risks, damages and disputes not being resolved until years of litigation between companies or regulators. I don't think there's much understanding of what it means to have. Companies then create legal precedents that may provide some clarity.
What is the best way to build AI responsibly?
The challenge in building AI responsibly is that many of the foundational pillars of responsible AI, such as fairness and safety, are based on normative values, and there is no common definition or understanding of these concepts. does not exist. Therefore, perhaps acting responsibly can still cause harm, and acting maliciously relies on the fact that there is no shared norm of these concepts to insist on acting with integrity. There is a possibility. Until there is a global standard or shared framework for building AI responsibly, the best way to pursue this goal is through clear principles, policies, and policies for the responsible development and use of AI. It's about setting guidance, standards and enforcing them through internal oversight. Benchmarking and other governance practices.
How can investors more effectively promote responsible AI?
Investors can do a better job of defining or at least clarifying what constitutes responsible AI development or use and taking action when AI actors’ practices are inconsistent. Currently, “responsible” or “trustworthy” AI has become a de facto marketing term, as there are currently no clear standards for evaluating the practices of AI actors. Initial regulations such as the EU AI Act will establish governance and oversight requirements, but investors are also looking to develop better practices that center human values and social benefits on AI actors. There are still areas where this could be encouraged. However, if investors are reluctant to act when there is evidence of inconsistency or bad actors, there is little incentive to adjust behavior or practices.