In 15 presentations in the style of Ted Mit Faculty recently discussed his pioneering research, which includes social, ethical and technical considerations and knowledge, each of which supported by seed grants established by Social and ethical obligations of calculations (Hearts), cross -sectional initiative Schwarzman College of Computing. . Call for the proposal Last summer has met with almost 70 applications. The Committee with representatives of each school MIT and university convened to choose the winning projects that received up to $ 100,000 financing.
“Hearts are involved in conducting progress at the intersection of computers, ethics and society. Seed subsidies are aimed at inflammation of bold, creative thinking around complex challenges and possibilities in this space,” said Nikos Triakis, co-founder of the hearts and JC Penney Management Professor. “Thanks to the ethics of the computing myth of the research symposium, we thought that it is important not only to show the width and depth of research that shape the future of ethical calculations, but inviting the community to participate in the conversation.”
“What you see here is a kind of common judgment of the community about the most exciting job when it comes to research, in the field of social and ethical duties related to calculation in myth,” said Caspar Hare, the co -existing dean of hearts and professor of philosophy.
. Full -day symposium On May 1, about four key topics were organized: responsible healthcare technology, artificial intelligence and ethics management, technology in society and civil involvement, as well as the inclusion of digital and social justice. The speakers provided presentations provoking thinking on a wide range of topics, including algorithmic prejudices, data privacy, social implications of artificial intelligence and evolving relations between people and machines. A poster session also appeared in the event, during which student researchers presented projects They worked all year round Scholars hearts.
The most important information from the myth of Ethics of Computing Research Symposium in each of the thematic areas, many of which are available to watch on YouTubeAttached:
Making the kidney transplant system fair
Politics regulating the organ transplantation in the United States are created by the National Committee, which often lasts over six months, and then years to implement, a schedule that many in the waiting list simply cannot survive.
Dimitris Betsimas, Vice President for Open Learning, Dean of Business Analytics and Operational Professor Boeing, He made his latest work available in the analytical allocation of honest and efficient allocation of kidney transplantation. The new Betsimas algorithm analyzes criteria such as geographical location, mortality and age in just 14 seconds, which is a monumental change from ordinary six hours.
Betsimas and his team work closely with United Network to sharing organs (UNOS), a non -profit organization that manages the majority of the national system of donations and transplants under a contract with the federal government. During his presentation, Betsimas released a film from James Alcorna, a senior political strategist of the Unos, which offered this poignant summary of the impact of the new algorithm:
“This optimization radically changes the duration of the assessment of these various simulations of political scenarios. It took us several months to look at a handful of different political scenarios, and now it takes the issue of minutes to review thousands and thousands of scenarios. We are able to introduce these changes much faster, which means that we can improve the system crossing candidates.
Ethics of social content generated by AI
Because the content generated by AI becomes more common on social platforms, what are the implications of disclosure (or not disclosure) that AI has been created any part of the post? Adam Berinsky, professor of political sciences Mitsui and dr Gabrielle Péloquin-Skulski in the Department of Political Sciences, studied this question at the session, which examined recent studies of the impact of various labels on the content generated by AI.
In a series of surveys and experiments placing labels for posts generated by AI, scientists looked at how specific words and descriptions influenced the perception of fraud by users, their intention to engage in post, and ultimately if the post was true or false.
“The great amount of our initial set of arrangements is that one size does not match everyone,” said Péloquin-Skulski. “We have found that the labeling of images generated by AI processed label reduces faith in both false and real posts. This is quite problematic, because labeling intends to reduce the faith of people in false information, not necessarily true information. It suggests that combining both process and truth can be better in counteracting Misinformation generated by AI.”
Using AI to increase civil discourse online
“Our research is aimed at solving how people are increasingly wanting to vote in organizations and communities to which they belong”, Lily Tsai explained during the session On experiments in generative artificial intelligence and the future of digital democracy. Tsai, professor of Ford Political Sciences and director Mit Governance Lab, conducts ongoing research with Alex Pentland, Professor Media Arts Science and a larger team.
Online platforms are recently gaining popularity in the United States in both the public and private sector environment. Tsai explained that thanks to technology everyone can have something to say – but it can be overwhelming and even feel dangerous. First of all, too much information is available, and secondly, online discourse is becoming more and more “non -cylinder”.
The group focuses on “how we can rely on existing technologies and improve them thanks to the rigorous, interdisciplinary research and how we can introduce innovations through integration of generative artificial intelligence in order to increase the benefits of online spaces.” They developed their own platform integrated by AI-Zintegrated by AI for deliberative democracy, meeting and implemented four initial modules. Until now, all research has been in the laboratory, but they are also working on the result of upcoming field research, the first of which they will participate in the government with the Colombia District government.
Tsai said to the audience: “If you don't take anything else from this presentation, I hope you will take it – that we should all demand technologies that are evaluated to see if they have positive lower level results, and not just focus on maximizing the number of users.”
Public Think Tank, which considers all aspects of AI
When Catherine d'Engazio, extraordinary professor of teaching and city planning, and Nikko Stevens, Postdoc in Data + Feminism Lab in myth, initially submitted their financing proposal, did not intend to develop a think tank, but the framework – the one that expressed artificial intelligence and machine learning could integrate social methods and use the project.
They finally created AI liberators, which they describe As “ongoing public Think Tank about all aspects of artificial intelligence.” D'Agnazio and Stevens gathered 25 researchers from various institutions and disciplines, who are the author of over 20 articles in positions, examining the most important academic literature on AI systems and commitment. They deliberately grouped the documents in three separate topics: corporate landscape AI, blind alleys and roads forward.
“Instead of waiting for open artificial intelligence or Google to invite us to take part in the development of their products, we gathered to challenge the status quo, think about a larger image and reorganization of resources in this system in the hope of greater social transformation,” said D'Agnazio.
