Uploaded April 2026 | Updated September 2026, 2 weeks ago
Machine Learning for Health (ML4H) Seminar Series
March 4, 2026
Talk Title: Escaping the AI Rut: Reimagining Learning, Purpose, and Power in Medicine
Speaker:
Leo Celi, M.D., M.Sc., MPH
MIT
Abstract:
We cannot escape the AI rut by doing more of the same—not with better algorithms, not with more data, not with incremental improvements to broken systems. The urgent re-imagination required spans how we engage with knowledge, how we learn in an age where lectures are obsolete, how we find purpose beyond efficiency metrics that benefit administrators over clinicians and patients, and how we operate within power structures that have medical professionals convinced they're smarter than everyone else while perpetuating the same health disparities year after year. MIT Critical Data has moved beyond benchmark datasets and technical workshops to fundamentally disrupt the status quo: organizing over 100 health data science events across 40 countries that deliberately interface people who don't think alike—bringing medical students together with community college students, indigenous knowledge holders, musicians, patients, and activists—because creativity and critical thinking emerge not from homogeneous classrooms but from challenging worldviews and exposing overlapping blind spots. Through our LLM-a-thons where actual users evaluate AI across languages and cultures, our Health AI Systems Thinking workshops developing policies universities and health systems can enact today, and our reimagined learning sessions integrating arts, indigenous philosophy, and religious traditions, we're teaching the next generation to push back against us—the elders maintaining the status quo—because solutions cannot come from those who preserve the problems, and youth engagement means nothing if they become complicit within a decade, trained to be exactly like us rather than empowered to change everything.
Speaker Bio:
Dr. Celi is the principal investigator behind the Medical Information Mart for Intensive Care (MIMIC) and its derivatives (MIMIC-CXR, MIMIC-ED, MIMIC-ECHO, and MIMIC-ECG), serving close to 100,000 users worldwide and generating over 10,000 publications that have fundamentally shaped machine learning in healthcare. His group at MIT leads MIT Critical Data, a global consortium whose mission is to build local AI capacity and agency in healthcare through open data and software, foster community engagement through data-thons and workshops, and advance equity through technology. His work pioneers process-based approaches to algorithmic fairness that move beyond traditional demographic metrics, developing culturally-aware human-AI systems that incorporate diverse knowledge traditions including indigenous epistemologies and faith perspectives. In partnership with hospitals, universities, and professional societies worldwide, Dr. Celi's team has organized over 100 health data science events across more than 40 countries, democratizing access to opportunities for growth and fostering interdisciplinary collaboration that actively addresses power dynamics in the global knowledge ecosystem.
For more information, visit: https://www.https://www.broadinstitute.org/ml4h
Copyright Broad Institute, 2026. All rights reserved.
Machine Learning for Health (ML4H) Seminar Series
March 4, 2026
Talk Title: Escaping the AI Rut: Reimagining Learning, Purpose, and Power in Medicine
Speaker:
Leo Celi, M.D., M.Sc., MPH
MIT
Abstract:
We cannot escape the AI rut by doing more of the same—not with better algorithms, not with more data, not with incremental improvements to broken systems. The urgent re-imagination required spans how we engage with knowledge, how we learn in an age where lectures are obsolete, how we find purpose beyond efficiency metrics that benefit administrators over clinicians and patients, and how we operate within power structures that have medical professionals convinced they're smarter than everyone else while perpetuating the same health disparities year after year. MIT Critical Data has moved beyond benchmark datasets and technical workshops to fundamentally disrupt the status quo: organizing over 100 health data science events across 40 countries that deliberately interface people who don't think alike—bringing medical students together with community college students, indigenous knowledge holders, musicians, patients, and activists—because creativity and critical thinking emerge not from homogeneous classrooms but from challenging worldviews and exposing overlapping blind spots. Through our LLM-a-thons where actual users evaluate AI across languages and cultures, our Health AI Systems Thinking workshops developing policies universities and health systems can enact today, and our reimagined learning sessions integrating arts, indigenous philosophy, and religious traditions, we're teaching the next generation to push back against us—the elders maintaining the status quo—because solutions cannot come from those who preserve the problems, and youth engagement means nothing if they become complicit within a decade, trained to be exactly like us rather than empowered to change everything.
Speaker Bio:
Dr. Celi is the principal investigator behind the Medical Information Mart for Intensive Care (MIMIC) and its derivatives (MIMIC-CXR, MIMIC-ED, MIMIC-ECHO, and MIMIC-ECG), serving close to 100,000 users worldwide and generating over 10,000 publications that have fundamentally shaped machine learning in healthcare. His group at MIT leads MIT Critical Data, a global consortium whose mission is to build local AI capacity and agency in healthcare through open data and software, foster community engagement through data-thons and workshops, and advance equity through technology. His work pioneers process-based approaches to algorithmic fairness that move beyond traditional demographic metrics, developing culturally-aware human-AI systems that incorporate diverse knowledge traditions including indigenous epistemologies and faith perspectives. In partnership with hospitals, universities, and professional societies worldwide, Dr. Celi's team has organized over 100 health data science events across more than 40 countries, democratizing access to opportunities for growth and fostering interdisciplinary collaboration that actively addresses power dynamics in the global knowledge ecosystem.
For more information, visit: https://www.https://www.broadinstitute.org/ml4h
Copyright Broad Institute, 2026. All rights reserved.
