AbstractIt is increasingly difficult to know who—or what—we are interacting with online. Traditional authentication methods prove credentials, not personhood, and AI-generated speech can now convincingly impersonate humans in real time. Autonomous agents can create accounts at scale, while deepfake detectors trained to spot artifacts remain brittle and often fail against new models, codecs, or channels. In high-stakes domains such as secure communications, defense operations, and financial approvals, this uncertainty poses significant risk. Without continuous assurance of human presence, systems remain vulnerable to spoofing.In this talk, I will present witness sensing: the use of auxiliary sensors physically and temporally coupled to media creation, providing hardware-rooted evidence of human origin. Unlike post-hoc detectors, witness sensing verifies authenticity at capture by comparing a primary modality (e.g., audio or video) with a secondary “witness” signal that reflects the same physical act but is difficult to spoof. As an example, I will focus on speech: our embedded prototype fuses radar and microphone sensing to capture articulatory motion, vocal-fold vibration, heartbeat, and respiration, synchronizing these signals with the acoustic waveform to continuously confirm both human presence and the embodied origin of speech. The result is a low-latency, tamper-resistant signal that persists throughout an interaction and remains robust to advances in generative AI.BiographyVisar Berisha is a professor at Arizona State University with a joint appointment in the College of Engineering and the College of Health Solutions, and also serves as Associate Dean for Research Commercialization in the College of Engineering. His research sits at the intersection of speech, AI, and human communication, with a broad focus on developing technologies that improve the human condition. These range from clinical applications of speech AI in healthcare to methods that protect against the risks of generative AI and ensure the authenticity of human communication. His work, supported by the National Institutes of Health, the Department of Defense, and the National Science Foundation, has led to many academic publications and patents, as well as two startups. Berisha’s research has been featured in the New York Times, ESPN, NPR, and the Wall Street Journal. He was the International Speech Communication Association's (ISCA) 2023–24 Distinguished Lecturer.

Speaker: Stefan Wager (Stanford University)Title: Learning to Price Electricity for Optimal Demand ResponseAbstract: The time at which renewable (e.g., solar or wind) energy resources produce electricity cannot generally be controlled. In many settings, however, consumers have some flexibility in their energy consumption needs, and there is growing interest in demandresponse programs that leverage this flexibility to shift energy consumption to better match renewable production — thus enabling more efficient utilization of these resources. We study optimal demand response in a setting where consumers use home energy management systems (HEMS) to autonomously adjust their electricity consumption. Our core assumption is that HEMS operationalize flexibility by querying the consumer for their preferences and computing the “indifference set” of all energy consumption profiles that can be used to satisfy these preferences. Then, given an indifference set, HEMS can respond to grid signals while guaranteeing user-defined comfort and functionality; e.g., if a consumer sets a temperature range, a HEMS can precool and preheat to align with peak renewable production, thus improving efficiency without sacrificing comfort. We show that while pricebased mechanisms are not generally optimal for demand response, they become asymptotically optimal in large markets under a mean-field limit. Furthermore, we show that optimal dynamic prices can be efficiently computed in large markets by only querying HEMS about their planned consumption under different price signals. We leverage this result to build an online contextual pricing algorithm, and show it to enable considerable reduction in peak system load in simulators calibrated to a number of major US cities.Biography: Stefan Wager is an associate professor of Operations, Information, and Technology at the Stanford Graduate School of Business, an associate professor of Statistics (by courtesy), and the Philip F. Maritz Faculty Scholar for 2025-26. His research lies at the intersection of causal inference, optimization, and statistical learning. He is particularly interested in developing new solutions to problems in statistics, economics and decision making that leverage recent advances in machine learning. He is currently serving as an associate editor for several publications including Biometrika, Management Science, Operations Research, and the Journal of the American Statistical Association. He has worked with or consulted for several Silicon Valley companies, including Dropbox, Facebook, Google, and Uber.

The Mobility Forum with Prof. Jinhua Zhao showcases transportation research and innovation across the globe. The Forum is online and open to the public.

Join us next Friday, October 24th, 12–1PM in 10-401 (Fish Bowl) for a talk with Brian Smith, Director of QLAB, and Alex Marshall, MIT MArch '13 alum and Vice President at Quarra Stone Company.Founded in 1989 with just six craftsmen, Quarra Stone Company has grown into one of the leading architectural stone fabricators in the U.S. Based in Madison, Wisconsin, Quarra bridges centuries-old craftsmanship with cutting-edge technology. From hand carving and 3D scanning to robotic milling and digital modeling.Their projects span from the U.S. Capitol to House of Horns by WOJR to contemporary art installations such as MIT’s own Officer Sean Collier Memorial, translating design intent into enduring material form. This work not only preserves stone’s ancient legacy but redefines its role in shaping today’s built environment.

*Special seminar: unusual time and locationSpeaker: Pierre Godfard (UNC Chapel Hill)Title: Rigidity of SU(2) quantum representations of mapping class groups at prime levelsAbstract:The property (T) conjecture for mapping class groups predicts that finite-dimensional unitary representations of mapping class groups of surfaces of genus at least 3 have no infinitesimal deformations (are rigid). This rigidity question has been extensively studied for finite image representations, where it is known as the Ivanov conjecture. A natural direction is to investigate infinite image unitary representations. Natural examples arise from unitary modular fusion categories via the Reshetikhin-Turaev construction, yielding what are called quantum representations of mapping class groups.For closed surfaces of genus at least 7, I will discuss rigidity of quantum representations arising from SU(2) and SO(3) modular categories at prime roots of unity. The strategy is to relate the problem to Ocneanu rigidity—a result asserting that modular fusion categories admit no non-trivial deformations. Ocneanu rigidity reduces our problem to constructing an injective map from deformations of the mapping class group representation in question to deformations of the modular fusion category. The construction uses modular functor theory and harmonic representatives from Hodge theory. Injectivity follows from a partial homological stability result for first cohomology groups of mapping class groups with coefficients in adjoints of SU(2) and SO(3) quantum representations.Zoom: https://mit.zoom.us/j/93615455445. For the Passcode, please contact Pavel Etingof at etingof@math.mit.edu.

Ten minutes for your mind@2:50 every day at 2:50 pm in multiple time zones:Europa@2:50, EET, Athens, Helsinki (UTC+2) (7:50 am EST) https://us02web.zoom.us/j/88298032734Atlantica@2:50, EST, New York, Toronto (UTC-4) https://us02web.zoom.us/j/85349851047Pacifica@2:50, PST, Los Angeles, Vancouver (UTC=7) (5:50 pm EST) https://us02web.zoom.us/j/85743543699Almost everything works better again if you unplug it for a bit, including your mind. Stop by and unplug. Get the benefits of mindfulness without the fuss.@2:50 meets at the same time every single day for ten minutes of quiet together.No pre-requisite, no registration needed.Visit the website to view all @2:50 time zones each day.at250.org or at250.mit.edu