Generative AI Innovation Incubator
Events
In summer 2023, we are offering a series of virtual events designed to engage a broad audience. Registration is required for all events, but there is no fee to register and attend. All events will be held virtually via Zoom. All events will be recorded and linked on this page for later viewing.
All attendees, sponsors, partners, volunteers and staff at our events are required to agree with our code of conduct. Organizers will enforce this code throughout the event. We expect cooperation from all participants to ensure a safe environment for everyone.
Use #GAI3 to connect with us on social media!
Upcoming Events
This summer's programming has ended, but we're working on adding additional events in the coming months. Watch this page for details!
Talk and Panel Series
Our talk series and panel series are offered to the broadest audience. These open events occur at 11 a.m. EDT on Fridays, and feature leading researchers and practitioners in the generative AI space, both generally and in our three impact areas: education and the future of work; medicine and public health; and finance and economics.
Tutorials
We will offer three generative AI tutorials to participants with computer programming and machine learning experience. The tutorials include lecture materials and hands-on practice.
Hackathons
A central focus of the summer 2023 events will be three extensive hackathons, each of which will comprise a series of events designed to foster community engagement around design and development of generative AI in three impact areas: education and the future of work; medicine and public health; and finance and economics.
Each of these hackathons will start with a panel discussion where experts working in the area will share their work on the frontier of generative AI and human impact. Registration for the hackathon will open when the panel discussion begins. To register, interested participants must declare an idea to contribute as well as what expertise they bring to a project. Upon registration, interested parties will be invited to join a Slack space where they will have the opportunity to meet and interact with other participants. Over subsequent weeks, participants will work together to offer each other feedback, refine ideas and ideate.
A mixer event will afford synchronous interaction and community-wide idea ranking. Subsequent to that event, individuals will form teams to work together on group proposals, which will also undergo a feedback and refinement process.
A few weeks prior to an intensive hackathon work weekend, four to six proposals will be selected to be supported for that event. At the conclusion of the three-day hackathon weekend, a panel will choose a winning team project. Project teams will be provided with their own private communication space, a GitHub space, a private chat space within Slack and compute resources on a SLURM cluster.
Winning teams will be offered financial support from the Block Center to further develop their project into an end-user application in the associated impact area, and will present their work at the closing ceremony.
Past Events
May 19, 2023 | 11 a.m.–1 p.m.
Watch VideoOPENING KEYNOTE TALK: Tom Mitchell
+ Mixer Activity
Tom Mitchell is the Founders University Professor at CMU, where he founded and chaired the world's first Machine Learning Department and served as interim dean from 2018 to 2019. His research interests include machine learning, artificial intelligence, cognitive neuroscience and the impact of AI on society. Mitchell is an elected member of the U.S. National Academy of Engineering and the American Academy of Arts and Sciences, and a fellow and past president of the Association for the Advancement of Artificial Intelligence (AAAI).
Mitchell has advised a variety of government, for-profit and nonprofit organizations, especially regarding their AI strategies and AI-related business opportunities. He has testified to a variety of U.S. congressional committees regarding potential uses and impacts of artificial intelligence. Mitchell currently co-chairs both a U.S. National Academies study on AI and the future of work and a task force for the Special Competitive Studies Project to study impacts of recent and future generative AI models on society and to recommend U.S. government responses.May 26, 2023 | 11 a.m.–1 p.m.
KEYNOTE TALK: Roni Rosenfeld
Roni Rosenfeld (B.Sc, mathematics and physics, Tel-Aviv University; Ph.D., computer science, CMU) is head of the Machine Learning Department and professor of machine learning, language technologies, computer science and computational biology in CMU's School of Computer Science. He has taught machine learning and statistical language modeling to thousands of undergraduate and graduate students since 1997, and has been a mentor to five post-doctoral students and an advisor to a dozen Ph.D. students and many master's and undergraduate students.
Roni’s current research interests are in tracking and forecasting epidemics. The Delphi research group, which he co-founded and has co-led since 2012, has been playing a leading role in developing epidemic forecasting technology in the U.S., and was named a National Center for Epidemic Forecasting by the U.S. CDC.
Roni has previously worked in statistical language modeling, speech recognition, human machine speech interfaces, and the use of speech and language technologies to aid international developments. He has published some 150 scientific articles in academic journals and peer reviewed conferences, and received the Spira Teaching Excellence Award (2017) and the Allen Newell Medal for Research Excellence (1992, 2022).
Ask Me Anything (AMA) Panel
Following the keynote talk, we will host a facilitated Ask Me Anything Panel with a set of AI experts ready to field questions related to capabilities, challenges, opportunities and caveats related to generative AI in the world.
Nicholas Mattei
Assistant Professor, Tulane University
Research areas: artificial intelligence, data science, decision-making, preferences
Biplav Srivastava
Professor, University of South Carolina
Research areas: artificial intelligence planning, learning and representation; smart cities, services
John Zimmerman
Tang Family Professor of Artificial Intelligence and Human-Computer Interaction, Carnegie Mellon University
Research areas: human-AI interaction, AI innovation process, human-robot/agent interaction, designing tech and policy simultaneously
June 2, 2023 | 11 a.m.–12:30 p.m.
PANEL: Generative AI in Education and the Future of Work
Researchers and practitioners in the learning sciences and technology, economics, and administration of education will talk about new challenges and opportunities resulting from the rise of large language models (LLMs). The discussion will focus on assessment, formal instruction, professional development and learning on the job. The changing employment landscape will also be addressed.
Sherice Clarke
University of California, San Diego
Shayan Doroudi
University of California, Irvine
Lewis Johnson
Alelo, Inc.
Nneka McGee
San Benito Consolidated Independent School District
Sherry Wu
Carnegie Mellon University
June 9, 2023 | 11 a.m.–12:30 p.m.
PANEL: Generative AI in Medicine and Public Health
Across government, industry and academia, LLMs are impacting medicine and public health in a variety of areas including diagnostic imaging and extraction of information from medical records. In this event, we bring together experts in computational biology, medical informatics and natural language processing to discuss privacy concerns, liability and trust in this important space.
Denis Newman-Griffis
University of Sheffield
Rada Mihalcea
University of Michigan
Aakanksha Naik
Allen Institute of Technology
Eric Xing
Carnegie Mellon University and Mohamed bin Zayed University of AI
Min Xu
Carnegie Mellon University
June 10, 2023 | 11 a.m.–2 p.m.
TUTORIAL: Building Blocks of Generative AI Systems and Hands-On Experience
Daphne Ippolito is an assistant professor in the Language Technologies Institute at CMU. Before joining the CMU community, she completed her Ph.D. at the University of Pennsylvania in LLMs, specifically in the area of language generation. She is also a research scientist at Google. Her interests include the properties of generative language models for text and the challenges of automatically generating narratives that are simultaneously coherent and interesting. She also looks at ways human writers can use generative text models as creative tools.
In the first half of the Generative AI tutorial, you will learn about the building blocks of modern neural language models and text-to-image diffusion models. We will go over the terminology commonly seen in technical discussions of these models, and we will describe step-by-step how they turn your input prompt into generated text or images. We will also discuss the key differences between popular models, and why you might choose one or another.
In the second half of the tutorial, you will complete a guided exercise to build a basic application using these models. The practical portion of the tutorial will assume competency in Python and access to a Python development environment, such as Google Colab.
June 16, 2023 | 11 a.m.–12:30 p.m.
PANEL: Finance and Economics
Opportunities for AI to impact finance and economics have grown in abundance over the recent past as the largest financial institutions have made an investment in these technologies with an eye toward transforming business processes. Leaders from industry and academia will discuss challenges related to process-optimization broadly, as well as quantitative reasoning, extrapolation and multimodal document processing in particular.
John Horton
Massachusetts Institute of Technology
Anton Korinek
University of Virginia
David Rosenberg
Bloomberg
Sameena Shah
JPMorgan Chase
Sean Welleck
AI2 and Carnegie Mellon University
June 18, 2023 | 11 a.m.–2 p.m.
TUTORIAL: Building Blocks of Generative AI Systems and Hands-On Experience
Daphne Ippolito is an assistant professor in the Language Technologies Institute at CMU. Before joining the CMU community, she completed her Ph.D. at the University of Pennsylvania in LLMs, specifically in the area of language generation. She is also a research scientist at Google. Her interests include the properties of generative language models for text and the challenges of automatically generating narratives that are simultaneously coherent and interesting. She also looks at ways human writers can use generative text models as creative tools.
In the first half of the Generative AI tutorial, you will learn about the building blocks of modern neural language models and text-to-image diffusion models. We will go over the terminology commonly seen in technical discussions of these models, and we will describe step-by-step how they turn your input prompt into generated text or images. We will also discuss the key differences between popular models, and why you might choose one or another.
In the second half of the tutorial, you will complete a guided exercise to build a basic application using these models. The practical portion of the tutorial will assume competency in Python and access to a Python development environment, such as Google Colab.
June 23, 2023 | 11 a.m.–12:30 p.m.
INVITED TALK: Eric Xing | From X-Ray Crystallography to AlphaFold to Generative AI: A ReAIssance of Empiricism and Connectionism in Biological Science
Eric Xing is a computer scientist, academic administrator, and entrepreneur. Prior to his appointment as president of the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), Xing was already a professor in the School of Computer Science at CMU and a researcher in machine learning, computational biology and statistical methodology. Eric is also the founder, chair, chief scientist and former CEO of Petuum Inc.
Eric's principal research interests lie in the development of machine learning and statistical methodology, and large-scale computational systems and architecture for solving problems involving automated learning, reasoning and decision-making in high-dimensional, multimodal and dynamic possible worlds in artificial, biological and social systems.
June 30, 2023 | 11 a.m.–12:30 p.m.
INVITED TALK: Manuela Veloso | AI in Finance: Examples and Discussion
In this talk, Manuela will present examples of recent AI research and practice experience in the finance domain, addressing data, reasoning and execution AI approaches. Presented projects will be on AI for data discovery, data standardization, synthetic data, behavior understanding, multiagent simulations and explainability.
Manuela Veloso is head of J.P. Morgan Chase AI Research and the Herbert A. Simon University Professor Emerita at CMU, where she was previously faculty in the Computer Science Department and head of the Machine Learning Department. Her recent interests are in AI, symbiotic human-robot autonomy, continuous learning systems and AI in finance. She is past president of the Association for the Advancement of Artificial Intelligence (AAAI), and the co-founder and a past president of the RoboCup Federation. She has received numerous awards and honors, including a National Science Foundation CAREER Award, the Allen Newell Medal for Excellence in Research, a Radcliffe Fellowship, the Einstein Chair Professor of the Chinese Academy of Sciences, and the ACM/SIGART Autonomous Agents Research Award. Veloso is a fellow of AAAI, AAAS, ACM and IEEE. In 2022, she was elected to the National Academy of Engineering for her “contributions to artificial intelligence and its applications in robotics and the financial service industry.”
July 14, 2023 | 11 a.m.–12 p.m.
INVITED TALK: Jill Fain Lehman | Separating Truth and (Speculative) Fiction
In his May GAI3 keynote, Roni Rosenfeld characterized the lesson of the 2010s as "Meaning lives comfortably in a Cartesian space," and the lesson of the current "Transformer Age" as "with good representation of meaning, you can do very impressive things." In this talk I will look more closely at what meaning is for human language, and what it isn’t for LLMs. The difference between the two suggests a way forward that respects human meaning while still allowing for a future in which we do very impressive things. In the speculative, cybernoir world of "Private I," my co-authors and I describe one such future, where the character of Marlowe offers an alternative view of symbiotic AI that is decidedly not based on today's statistical approach to language interaction.
Jill Fain Lehman (B.S., Yale, computer science and psychology; Ph.D., CMU, computer science) pursues research at the intersection of AI, cognitive science and language. Her thesis work on adaptive parsing was one of the early approaches to combining NLP with reactive, incremental machine learning, and she later contributed an initial theory of language to the Soar cognitive architecture. In industry, she has worked on language-based interaction for the Rand Corporation, Carnegie Speech and Disney Research, and is the former owner of KidAccess Inc., a company that provided cognitively based communication materials and training for young children with autism and their service providers. She is currently a senior project scientist in the Human-Computer Interaction Institute at CMU.
July 18, 2023 | 11 a.m.–2 p.m.
TUTORIAL: Building Blocks of Generative AI Systems and Hands-On Experience
Daphne Ippolito is an assistant professor in the Language Technologies Institute at CMU. Before joining the CMU community, she completed her Ph.D. at the University of Pennsylvania in LLMs, specifically in the area of language generation. She is also a research scientist at Google. Her interests include the properties of generative language models for text and the challenges of automatically generating narratives that are simultaneously coherent and interesting. She also looks at ways human writers can use generative text models as creative tools.
In the first half of the Generative AI tutorial, you will learn about the building blocks of modern neural language models and text-to-image diffusion models. We will go over the terminology commonly seen in technical discussions of these models, and we will describe step-by-step how they turn your input prompt into generated text or images. We will also discuss the key differences between popular models, and why you might choose one or another.
In the second half of the tutorial, you will complete a guided exercise to build a basic application using these models. The practical portion of the tutorial will assume competency in Python and access to a Python development environment, such as Google Colab.
July 21–23, 2023 | 11 a.m.–8 p.m.
HACKATHON EVENT: Education and the Future of Work
Participants with some technical expertise who would like to engage in an extensive hands-on group project in one of our three impact areas can join one of three hackathon experiences. These events will culminate in an open demonstration of hackathon projects and distribution of prizes to winning teams.
July 22, 2023 | 12–1 p.m.
SPOTLIGHT SESSION: Generative AI in Education and the Future of Work
Each of our hackathon events will feature a spotlight session, where ongoing CMU projects related to the thematic area of the hackathon will present their work in a poster/demo/firehose talk session.
Generative AI and Writing Instruction
Suguru Ishizaki, Professor, Department of English, Carnegie Mellon University
David Kaufer, Professor Emeritus, Department of English, Carnegie Mellon University
Nick Ryan, Director of Strategic Initiatives, Department of English, Carnegie Mellon University
Neural Model Interpretability in Writing Assessment
James Fiacco, Ph.D. student, Language Technologies Institute, Carnegie Mellon University
July 28–30, 2023 | 11 a.m.–8 p.m.
HACKATHON EVENT: Medicine and Public Health
Participants with some technical expertise who would like to engage in an extensive hands-on group project in one of our three impact areas can join one of three hackathon experiences. These events will culminate in an open demonstration of hackathon projects and distribution of prizes to winning teams.
July 29, 2023 | 12–1 p.m.
SPOTLIGHT SESSION: Generative AI in Medicine and Public Health
Each of our hackathon events will feature a spotlight session, where ongoing CMU projects related to the thematic area of the hackathon will present their work in a poster/demo/firehose talk session.
AI and Computational Biology
Robert F. Murphy, Ray and Stephanie Lane Professor Emeritus, Computational Biology Department, Carnegie Mellon University
AI, Healthcare and the CMLH/CIH
Ari Lightman, Distinguished Service Professor, Heinz College, Machine Learning Departmen and Center for Machine Learning and Health, Carnegie Mellon University
Aug. 4–6, 2023 | 11 a.m.–8 p.m.
HACKATHON EVENT: Finance and Economics
Participants with some technical expertise who would like to engage in an extensive hands-on group project in one of our three impact areas can join one of three hackathon experiences. These events will culminate in an open demonstration of hackathon projects and distribution of prizes to winning teams.
Aug. 5, 2023 | 12 p.m.–1 p.m.
SPOTLIGHT SESSION: Generative AI in Finance and Economics
Each of our hackathon events will feature a spotlight session, where ongoing CMU projects related to the thematic area of the hackathon will present their work in a poster/demo/firehose talk session.
FinQA: Testing Numerical Reasoning in LLMs
Anusha Rao, Master's Student in Computational Data Science, Carnegie Mellon University
Siddharth Parekh, Undergraduate in Computer Science, Carnegie Mellon University
Generative AI and the Mobile Market
Tianqi Chen, Assistant Professor, Machine Learning Department, Carnegie Mellon University
Generative AI and Privacy
Norman Sadeh, Professor, Software and Societal Systems Department, Carnegie Mellon University
Aug. 11, 2023 | 11 a.m.
Hackathon Demo Session | Closing Ceremony | Awards
Each of our hackathon events will include a competition for Best Project to support continued work on the hackathon project of developing an end user application. Awards will be provided by the Block Center. A closing session will offer the opportunity for the winning project teams to present their award-winning work to the general public. Audiences will have the chance to ask questions and engage in a broader discussion about the future of generative AI during this facilitated event.
Featured Keynote Speakers:
Mona Diab is lead Responsible AI scientist at Meta, and will be stepping into a new role as director of CMU's Language Technologies Institute. She is famous for her work on responsible and Arabic natural language processing, and is also generally passionate about language, mind, technology/society, history, politics and nutrition.