There has been a significant surge in the use of algorithms and artificial intelligence to address various problems and challenges. The rise of AI in particular is reshaping nearly every industry, profession, and field of study, but such innovations can have unintended consequences with new norms, new expectations, and new rules and laws. is common.
To foster deeper understanding, the Social and Ethical Responsibility of Computing (SERC), a cross-cutting initiative of the MIT Schwarzman School of Computing, recently launched a collaboration between social scientists and humanists, computer scientists, and engineers. , and other computing faculty, to demonstrate how the wide applicability of algorithms and AI has brought both opportunities and challenges to many aspects of society.
“The very nature of our reality is changing. AI is disrupting what until recently was the realm of human intelligence: our understanding of what it means to be human.” We have the ability to do what is possible,” said Daniel Huttenrocker, dean of the MIT Schwarzman School of Computing, in his opening remarks. At the 1st SERC Symposium. “This raises philosophical, conceptual, and practical questions on a scale not experienced since the beginning of the Enlightenment. In the face of such profound change, we must change We need a new conceptual map to get us through.”
The symposium provided a glimpse into SERC’s vision and activities in both research and education. Georgia Perakis, William F. Pounds Professor of Management at the MIT Sloan School of Management, said, “Our responsibility to SERC is to educate and equip our students and ensure that our faculty contribute to the responsible development and deployment of technology. I believe that we can do this.” He is SERC’s associate dean and symposium organizer. “We are leveraging the many strengths and diversity of disciplines within and outside of MIT and integrating them to gain multiple perspectives.”
The symposium delved into a variety of topics related to social and ethical aspects of computing through a series of panels and sessions. In addition, 37 undergraduate and graduate students from a variety of majors, including urban studies and urban planning, political science, mathematics, biology, electrical engineering and computer science, and brain and cognitive science, participated in a poster session and discussed the research. I published my research. In the space, he covers topics such as quantum ethics, AI collusion in the storage market, computing waste, and empowering users on his social platform to increase the trustworthiness of content.
Introducing a variety of works
The SERC Symposium featured research work by 12 faculty across these areas in three sessions dedicated to the themes of useful and just computing, fair and personalized health, and algorithms and humans.
One such project, by a multidisciplinary team of archaeologists, architects, digital artists, and computational social scientists, aimed to preserve Afghanistan’s endangered heritage sites in digital twins. The project team created highly detailed, interrogable 3D models of the ruins, along with augmented reality and virtual reality experiences, as learning resources for audiences who don’t have access to these ruins.
In the Unified Network for Organ Sharing project, researchers are conducting an applied analysis of the United States’ organ allocation system, which is currently undergoing a major overhaul to make it more efficient, fair, and inclusive. We have shown how to use it to optimize various aspects. For different races, ages, gender groups, etc.
Another talk discussed the broader equity implications of biased sensor data for next-generation models in computing and healthcare, areas that have not yet received enough public attention.
The Algorithmic Bias talk looked at both human bias and algorithmic bias, and the potential to improve results by considering the different nature of the two types of bias.
Other notable research includes the interaction of online platforms and human psychology. A study of whether decision makers make systematic prediction errors regarding available information. You will also learn how advanced analytics and calculations can be leveraged to inform supply chain management, operations, and regulatory operations in the food and pharmaceutical industries.
Improved tomorro algorithmw
“Algorithms undoubtedly impact every aspect of our lives,” said Dr. Schwarzman, associate dean and head of the Department of Electrical Engineering and Computer Science in MIT’s Schwarzman College of Computing. said one As Ozdagler at the start of a panel discussion he moderated. Impact of data and algorithms.
“There is no doubt that there will be much more to come, whether it be in the context of social media, online commerce, automated tasks, or even broader creative interactions with the advent of generative AI tools and large-scale language models. No,” Ozdagler said. . “While its promise is clear to all of us, there is also much to be concerned about. Now is a time for imaginative thinking and careful deliberation to improve tomorrow’s algorithms.”
Ozdagler turned to the panel and asked experts in computing, social science, and data science about how to understand what is coming and how to shape it to deliver enriching outcomes for the majority of humanity. I asked for insight.
Sarah Williams, associate professor of technology and urban planning at MIT, emphasized that data is the foundation of all models, so understanding the process of how datasets are assembled is critical. did. She also emphasized the need for research to address the potential impact of algorithmic bias, which often creeps in through the creators of the algorithms and the data used to develop them. “It’s up to us to come up with our own ethical solutions to these problems,” she says. “As important as advances in technology are, we need to start the field looking at the question of what biases are there in the algorithms? What biases are there in the data, or the process of that data?” is there?”
The panelists shifted their focus to generative models and whether the development and use of these technologies should be regulated, and included MIT’s Srini Devadas Professor of Electrical Engineering and Computer Science, John Houghton Professor of Information Technology, and Simon Johnson. Professors were also included. Entrepreneurship — Everyone agrees that it is difficult to regulate publicly accessible open source algorithms, given that regulators are still playing catch-up and have difficulty even putting guardrails on 20-year-old technology. matched.
Returning to the question of how to effectively regulate the use of these technologies, Mr. Johnson proposed a progressive corporate tax system as a potential solution. He recommends that companies pay taxes based on profits, especially large companies whose vast profits are mostly tax-free through offshore banking. In doing so, the approach could serve as a regulatory mechanism to stop companies from trying to “own the whole world” by imposing disincentives, Johnson said.
The role of ethics in computing education
With advances in computing showing no signs of slowing down, it is important to educate students to intentionally consider the social implications of the technologies they develop and deploy around the world. But can such things actually be taught? If so, how?
Casper Hare, professor of philosophy at MIT and co-associate dean of SERC, asked faculty this pressing question during a panel he moderated on the role of ethics in computing education. Each panelist, all with experience teaching ethics or thinking about the social impact of computing, shared their perspectives and approaches.
Eden Medina, associate professor of science, technology, and society at the Massachusetts Institute of Technology, is a strong advocate of the importance of learning from history. One of the things I do in my teaching is look at how people have faced these problems in the past and try to draw from that as a way to think about possible ways forward. . ” Medina regularly uses case studies in his classes, referencing a paper written by Joanna Radin, a historian of science at Yale University, on the Pima Indian diabetes dataset. This paper raises ethical questions about a particular history of data collection that many do not consider as examples. How decisions about technology and data evolve from very specific situations.
Milo Phillips Brown, Associate Professor of Philosophy at the University of Oxford, spoke about the Ethical Computing Protocol he co-created during his time as a SERC postdoc at MIT. The protocol is his four-step approach to building technology responsibly, which helps computer science students better understand and understand the social impact of technology by breaking down the process into more manageable steps. Designed to help you think accurately. “The fundamental approach we take draws heavily on the fields of value-based design, responsible research and innovation, and participatory design as guiding insights, and is also fundamentally interdisciplinary.” he said.
Fields such as biomedicine and law have ethical ecosystems that distribute the functions of ethical reasoning across these fields. It provides oversight and regulation to guide front-line stakeholders and decision makers when issues arise, as well as access to training programs and multidisciplinary expertise. “There’s nothing like that in this field,” says John Bustle, an associate professor of philosophy at Northeastern University. “With the current generation of computer scientists and other decision-makers, we’re actually letting them make their own ethical reasoning,” Basl added, adding that in philosophy classes alone, It is essential that core ethical reasoning skills are taught throughout the curriculum, and the goal is not that all computer scientists become professional ethicists, but that enough computer scientists become professional ethicists. He commented that it is about acquiring knowledge. You can ask the right questions and look for relevant expertise and resources that exist.
After the final session, an interdisciplinary group of faculty, students, and researchers engaged in a lively discussion on the issues addressed throughout the day at a reception that concluded the symposium.