Alvaro Velasquez, PHD

Program Manager
Defense Advanced Research Projects Agency (DARPA)

Alvaro Velasquez is a program manager in the Innovation Information Office (I2O) of the Defense Advanced Research Projects Agency (DARPA), where he currently leads the Assured Neuro-Symbolic Learning and Reasoning (ANSR) program. Before that, Alvaro oversaw the machine intelligence portfolio of investments for the Information Directorate of the Air Force Research Laboratory (AFRL). Alvaro received his PhD in Computer Science from the University of Central Florida and holds an interdisciplinary research record, including publications in artificial intelligence, combinatorial optimization, networking, cloud computing, and logic and circuit design. Alvaro is a recipient of numerous awards, including the National Science Foundation Graduate Research Fellowship Program (NSF GRFP) award, the University of Central Florida 30 Under 30 award, and best paper and patent awards from AFRL. He serves as Associate Editor of IEEE Transactions on Artificial Intelligence and his research has been funded by the Air Force Office of Scientific Research.

ABSTRACT

Challenges and Opportunities in Neuro-Symbolic Artificial Intelligence

Neuro-Symbolic Artificial Intelligence has experienced a renaissance and gained much traction in recent years as a potential “third wave” of AI to follow the tremendously successful second wave underpinned by statistical deep learning. This seeks the integration of neural learning systems and formal symbolic reasoning for more efficient, robust, and explainable AI. Such an integration holds much promise in areas like reinforcement learning and planning, where tremendous progress has been made in recent years, including great feats like the defeat of the world Go champion and powerful agents for real-time strategy games. However, the tremendous success of autonomous decision-making has highlighted its own shortcomings when it comes to data limitations, robustness, and trust, among other things. This talk presents some of these challenges and opportunities facing the development of neurosymbolic autonomy, how this differs from conventional neurosymbolic AI problems like classification and natural language processing, and potential implications to facilitating the broader adoption of autonomous solutions.