Welcome to my website. I am currently an Associate Professor of Machine Learning at Carnegie Mellon University (CMU). I hold appointments in the Machine Learning Department in the School of Computer Science, the Heinz School of Public Policy (courtesy) and Societal Computing (courtesy). My research spans core ML methods and theory, their applications in healthcare and natural language processing, and critical concerns, both about the mode of inquiry itself, and the impact of the technology it produces on social systems.
I am also the CTO and Chief Scientist of Abridge, a healthcare AI company defining the cutting edge of technology in the emerging ambient listening space. Our industry-leading product turns turns raw audio of doctor-patient conversations into high-quality drafts of after-visit documentation, freeing up doctors to focus on the patient (during the visit) and to focus mostly on last-mile edits (after the visit). Bringing this vision to life takes scientific innovation on every component of the technical stack, a commitment to focus on problem solving above aesthetic concerns, and to innovate in how to evaluate, monitor, and adapt models in light of changing patterns of usage.
I completed my PhD at the loveliest of universities (UCSD) in the Artificial Intelligence Group, and if I had a time machine, I would go back, take two years longer to graduate, and actually learn to surf.
At CMU, I direct the Approximately Correct Machine Intelligence (ACMI) Lab, a group of wonderful students whose creativity and talent are the primary reasons why my perennial threats to relocate to a small island in the Aegean Sea remain idle jests. Once I can no longer attract such talented kids to inspire me, I plan to abscond to Amorgos, where I will passively supervise a herd of goats, slowly acquire the centuries-old craft of distilling spirits from local herbs, and devote the rest of my life to writing third-rate science fiction novels.
My lab’s focuses include (i) building robust systems that can cope with a changing world, whether due to natural changes in the environment (so-called natural distribution shifts) or due to the strategic manipulations of other agents keen to influence automated decisions; (ii) understanding the social impacts of machine learning in a philosophically coherent way; (iii) the intersection of representation learning and causality; and (iv) leveraging ML to address impactful questions in clinical medicine. Across these many concerns, I tend to favor applications involving natural language data, and expect this interest to endure.
I value clear scientific prose and have (co-)authored two reviews of the literature (on RNNs and Differential Privacy), and more recently created an interactive book, (Dive into Deep Learning)[https://d2l.ai], which teaches deep learning through exposition, math and code, in a fully-interactive textbook written in Jupyter and automatically compiled to HTML and PDF (forthcoming on Cambridge University Press). In Fall 2016, I launched Approximately Correct, a blog aimed at bridging technical and social perspectives on machine learning. We have had some success addressing misconceptions about AI, both in the broader discourse and within the research community, but words can only do so much in the face of illiteracy.
For now, I plan to keep this site static, referring visitors to ACMI lab’s website for dynamically update content, including recent papers, current students, and other news.
Contact: zlipton [at] cmu [dot] edu.
Office: GHC 8129
Assistant: To maximize the likelihood that your email receives a response, please CC my administrative assistant, Marlee Bandish: mbandish [at] andrew [dot] cmu [dot] edu, whose organizational skills compensate for my lack.