Lucas is a research scientist at Numenta, a research company focused on neocortical theory, where he works on bringing neocortical principals to machine intelligence. Previously, Lucas worked as a data scientist for the Brazilian anticorruption agency helping to identify and prevent corruption schemes through the use of machine learning. He also founded Derivada, a non-profit organization that focuses on promoting AI through open source projects, research and education. Lucas brings his passion for building a collaborative AI community to Numenta where he created and leads the Brains@Bay Meetup group focused on brain-inspired machine learning algorithms.
Neuroscience-Driven AI: The Benefits of Sparsity for Artificial Intelligence
Today’s AI systems work in very different ways than the brain – consuming large amounts of power and relying on dense representations. The brain, on the other hand, is highly efficient – storing and processing information in a highly sparse manner. In this talk, I will discuss how sparsity in the brain can be applied to practical AI systems to improve robustness, continuous learning and computational efficiency. I’ll walk through an example of how Numenta’s sparse networks unlocked groundbreaking performance gains of more than 50x in deep learning networks with no loss of accuracy. I’ll then discuss how sparsity is just the beginning of a neuroscience-based approach that provides the path to general intelligence. By applying a brain-based theory to AI, we can address today’s bottlenecks while enabling tomorrow’s applications.