MIT Washington Seminar Series 2023-2024
Join the MIT Club of Washington’s 41st annual seminar series on an important national topic in science, technology, and public policy. Each year, the series offers engineers, scientists, industry leaders, policymakers and educators the opportunity to explore a specific topic in depth. Take this opportunity to better understand recent developments and key issues.
Each session begins at 6:15 p.m. with a cash bar, followed by dinner at 7:00 p.m. and a presentation from approximately 8:00 a.m. to 9:30 a.m. Presentations from distinguished speakers are followed by ample time for questions and discussions. The social hour and dinner provide additional opportunities to meet speakers, renew acquaintances, or engage in stimulating discussions with other attendees.
The series registration fee is $275 for ColumbiaDC members, which includes all dinners. The registration fee for non-members is $305.
Description of the 2023-2024 seminar series:
Artificial intelligence (AI) methods and applications have evolved significantly in recent years, notably with the November 2022 release of a new generative AI tool called ChatGPT. These current AI tools can already generate credible images, create documents in near-perfect English, reproduce personal voices and predict likely actions – and future improvements can be expected. But they also raise crucial policy questions such as accuracy, plagiarism, accountability and economic disruption. This seminar series will examine the methods and impacts of these emerging AI tools, as well as the ethical challenges they raise.
Session descriptions:
Tuesday November 14, 2023: Introduction to Contemporary AI
Speaker: Prof. Eric Burger ’84, research director, Commonwealth Cyber Initiative, Virginia Tech
The world of AI has progressed a lot in the last few years, even in the last few months. This includes machine learning, deep learning, neural networks, natural language models and now generative AI business models that predict with surprising accuracy what the user will do or say next. Google, Microsoft, and most other companies are already rapidly adding generative AI capabilities to their search and related products. But there are risks to using these new technologies, and we need to understand them. This introductory session will describe this evolution of AI, identify the capabilities of current AI tools, and how AI components relate to each other.
Tuesday December 19, 2023: Natural language models
Speaker: Prof. Jacob Andreas, Computer Science and Artificial Intelligence Laboratory (CSAIL), MIT
Natural language models are at the heart of many recent AI advances, including ChatGPT, which was released to the public in November 2022. Their capabilities allow users to interact in conversational English and generate results as text relevant and meaningful. Applications exist in areas such as chatbots, content creation, language translation, and text-to-speech synthesis. We need to know how large language models (LLMs) work and why they sometimes generate misleading or false results. This session will discuss how LLMs are created/used and explore the connection between large language models and generative AI, including examining neural networks, training algorithms, and the importance of effective prompts.
Tuesday January 9, 2024: The risks and possible regulation of using advanced AI tools
Speaker: Dr. Paul Werbos, former Program Director, Information Technology, NSF
AI models that predict future events based on digesting large amounts of data covering past events present several significant risks. Fairness, bias and transparency are key examples. Some argue that advances in AI could cause serious damage to our society, even its extinction. Current generative AI models make errors and sometimes produce incorrect or misleading predictions. As a result, there are calls for industry guidelines or government regulation on the rapid use of AI models – or at least a pause in the further development of advanced AI tools. This session will explore the validity and status of these risks as well as the push for regulation from a key player in funding AI research for decades.
Tuesday February 6, 2024: Future directions of AI research and applications
Speaker: Dr. Daniel Weld ’88, Professor Emeritus, Univ. of Washington, Managing Director and Chief Scientist of Semantic Scholar, Allen Institute of AI.
The rapid revolution in AI tools and applications leaves us wondering what comes next. The speaker will draw on his long career as an IT educator and entrepreneur to explore potential directions for AI. Where will the innovations be concentrated: in healthcare, education, cybersecurity, improving productivity? It will integrate the work of the Allen Institute for AI, a large Seattle-based nonprofit research organization whose mission is to conduct high-impact AI research and engineering for the greater good.
Tuesday March 19, 2024: Machine learning for water and agriculture
Speaker: Professor Feras Batarseh, Department of Biological Systems Engineering, Virginia Tech
A key part of the evolution of AI capabilities is the development and use of advanced machine learning and neural networks in different fields. This session will describe the steps necessary to configure and run an AI application using contemporary machine learning tools for agricultural and water systems. A recent application for water systems in the DC area will serve as an example. It will also explore the current competitive rush for many companies and supply chain producers to offer AI services and integrate generative AI tools into their products.
Tuesday April 9, 2024: The User Perspective: Exploring New AI Tools
Speaker: Joanna Stern, senior technology reporter, Wall Street Journal
Join us for an enlightening talk from Joanna Stern as she dives into the fascinating world of generative AI tools from a user perspective. Generative AI has revolutionized various industries, from art and design to language processing and healthcare. However, understanding and navigating these cutting-edge tools can be a complex journey for users. The speaker will draw on her extensive interactions with vendors and users to highlight the challenges and opportunities users face when using these new AI tools. She will explain how these tools can be used to improve our daily lives and what considerations are crucial for designing user-centered experiences. She will also talk about the main players in the AI industry and the directions this new AI technology could take.
Metro: Maggiano’s is located very close to the Friendships Heights subway station.
Car park: Discounted parking is available in the underground garage (marked Pavilion) for up to 4 hours – entrance is via Military Road and park in the section labeled “Friendship Center” (look for the elevator with the Maggiano sign) . Bring your parking ticket to Maggiano for validation.
The restaurant is accessible to people with disabilities.
