As a theoretical cryptographer, Kalai deals with questions like: how do you prove you know something without revealing what it is? What does it mean to know something at all? How can we verify AI answers are good, not just correct? And, in a world increasingly run by systems we don't fully understand, what do we owe each other in the name of safety?
Despite millions of years of evolution, the development of the human fetus remains a complex process that is vulnerable to genetic, environmental, and metabolic disruptions. These disruptions could result in developmental abnormalities, starting with one of the first organs to begin developing in the fetus that continues to develop after birth: the brain.
As artificial intelligence becomes more and more powerful, its ability to handle human-like tasks is undeniable. Artificial neural networks can learn to recognize images, use language, or write computer code. Like the brain’s circuits, they reorganize themselves as they learn, adjusting the strengths of their connections to improve their performance.
Please join this TechTalk from Google, a CSAIL Alliances member
The rapid evolution of the optical layer has profoundly shaped and differentiated compute infrastructure. Optical interconnects have been crucial in supporting the immense growth of networking bandwidth for machine learning, successfully scaling with electrical I/O while simultaneously improving power efficiency and cost. Consequently, machine learning networks have become the primary driver for business and technological advancements in fiber optic communications.
Add to calendarAmerica/New_YorkGoogle Tech Talk: Cross-layer Optimization for Scalable ML Networks11/04/2026
Please join this TechTalk from Google, a CSAIL Alliances member
The rapid evolution of the optical layer has profoundly shaped and differentiated compute infrastructure. Optical interconnects have been crucial in supporting the immense growth of networking bandwidth for machine learning, successfully scaling with electrical I/O while simultaneously improving power efficiency and cost. Consequently, machine learning networks have become the primary driver for business and technological advancements in fiber optic communications.
This presentation will explore various network architectures and technological considerations involved in building networks for emerging machine learning applications.
A particular focus will be placed on the roles of optical interconnects and optical switching, discussing relevant technology trends, key metrics, and a roadmap for the next 3 to 5 years, including the need for and challenges with tighter integration of optics and electronics.
In the most recent episode of Building 32, Popescu explains how her research sits at the intersection of computational design, digital fabrication, and structural engineering.
American Family Insurance, CSAIL Alliances Affiliate, exists to protect what matters most to households and businesses across the country, using data science and applied AI.
Add to calendarAmerica/New_YorkAI, Data Science & Careers from American Family Insurance10/15/2026
American Family Insurance, CSAIL Alliances Affiliate, exists to protect what matters most to households and businesses across the country, using data science and applied AI.
This talk introduces MIT graduate students to insurance as a serious technology, data, and AI problem, not just a financial product. Insurance AI rarely means one model: many specialized systems — for risk, fraud, compliance, and claims — each with its own evaluation, quality gates, and human review. The speaker leads several applied AI initiatives at American Family and has worked hands-on across this ecosystem. The talk draws on that complexity, and on pressures like catastrophe risk, climate, cybersecurity, and responsible AI deployment, to explain why insurance sits at the center of some of today's hardest technology problems. It highlights AmFam's academic research partnerships — on severe weather and wildfire risk — as one model for how graduate research translates into real-world impact, and closes with internship and career opportunities for graduate students.
Tesla, CSAIL Alliances members, develops and deploys autonomy at scale in vehicles, robots and more.
Tesla believes that an approach based on advanced AI for vision and planning, supported by efficient use of inference hardware, is the only way to achieve a general solution for full self-driving, bi-pedal robotics and beyond.
Add to calendarAmerica/New_YorkTraining Robust Optimus Policies: Tesla AI Tech Talk10/14/2026
Tesla, CSAIL Alliances members, develops and deploys autonomy at scale in vehicles, robots and more.
Tesla believes that an approach based on advanced AI for vision and planning, supported by efficient use of inference hardware, is the only way to achieve a general solution for full self-driving, bi-pedal robotics and beyond.
Tesla Full Self Driving (FSD) drives millions of miles autonomously every day across hundreds of cities. This feat is only possible through training extremely robust policies. When the goal is to similarly have millions of robots doing thousands of useful tasks in a wide variety of environments, we need to ensure that Optimus has similarly robust policies. In this talk, we share some of our strategies in training such robust policies and showcase some of our latest world models that make this possible.
MIT CSAIL researchers are joining the National Science Foundation’s (NSF) new Center for Human and Robot Co-Adaptation, where they’ll tackle a crucial question: What happens when assistive robots actually work with us?
