AI capabilities, and particularly machine learning (ML), crossed a scaling threshold a decade or so ago, and have since become a significant source of capability in a wide range of systems and in diverse sectors. In this talk, building on perspective granted from sixty years of DARPA investment in AI capabilities, we review the progress and prospects for AI-based systems, with emphasis on potential pathways to more trustworthy and reliable AI-based systems.

Enjoy a replay of The DEVOPS Conference sessions conveniently through the DevOps Sauna podcast. This talk is hosted by Dr. William Scherlis, Office Director at Information Innovation Office (I2O), DARPA