Episode 31: Science 2.0

  In this thematic episode of the Voices from DARPA podcast, three program managers discuss the possibility that emerging technologies in the arena of artificial intelligence (AI) are converging toward an “artificial-science” toolset that could open an era we might designate as Science 2.0. The prospect of AI scientists making Nobel-prize-caliber discoveries is not around the corner, but it is a distinct possibility for the future, suggests program manager Jiangying Zhou of the agency’s Defense Sciences Office (DSO). On the way toward that ideal, adds program manager Joshua Elliott of the Information Innovation Office (I2O), we are likely to rely on scientifically-minded AI tools to pump up the efficiency of scientific discovery and to tap into the vast and growing reservoirs of data, which biological minds might not be as suited to make sense of as AI ones. For Bartlett Russell, also of DSO, perhaps the most important advance during the evolution toward a Science 2.0 era will reside in the use of AI tools that enable more people than ever to embrace the scientific enterprise. The more minds doing science, she says, the more discovery we can expect.