Sarah Laszlo, PhD

Dr. Sarah Laszlo holds a bachelor's degree in Brain and Cognitive Science from MIT, where she was Phi Beta Kappa and winner of the Hans Lukas Teuber award for scholarly excellence. She completed her master's and doctoral degree in Brain and Cognitive science at the University of Illinois, Urbana-Champaign. Subsequently she was an NIH-NRSA postdoctoral fellow at Carnegie Mellon University for two years, which is where she began conducting machine learning research, as one of the very first practictioners to build what are now known as "deep learning" models. Sarah was a tenured professor of psychology and linguistics at Binghamton University, where her courses included artificial intelligence, graduate statistics, and cognitive architectures. Her computational linguistics work at Binghamton received an NSF CAREER award. At Google (X and Research) Sarah worked on reservoir and neuromorphic computing (including peta-scale computing), machine vision, and text-to-image generation (with an emphasis on safety and fairness). Models Sarah worked on in Google Research include Imagen, Bard, and Gemini. Additionally, Sarah was the technical lead of the team that contributed the reservoir computing module to Keras. She has 18 issued and 35 pending patents in the areas of bioamplification and reservoir computing from her time at Google. At Visa, Sarah owned the Machine Learning Platform, serving Visa’s corps of data scientists and optimizing Visa’s high performance computing resources, including an NVIDIA DGX SuperPOD. Sarah and her research have been featured on NPR’s Science Friday, as well as Wired, Scientific American, and the Huffington Post. In 2023, Sarah was made a Fellow of the Society for Psychophysiological Research, reflecting her career-long excellence in the psychophysical sciences. Sarah is available for consulting on AI safety and fairness, speaking engagements, and opportunities to mentor students.