Defense Technology Lecture Series: Artificial Intelligence for National Security Applications
The 2017 Belfer Center Study, Artificial Intelligence and National Security asserts that, “Existing capabilities in AI have significant potential for national security.” The report goes on to say, “Future progress in AI has the potential to be a transformative national security technology, on a par with nuclear weapons, aircraft, computers, and biotech…we argue future progress will be at least equally impactful.” This potential notwithstanding, current AI technologies still have their shortcomings. For example, deep machine learning requires huge amounts of training data to perform well, has problems generalizing, and is limited in its ability to handle common-sense reasoning and contextual constraints. It is difficult to project how far AI will advance in the next few decades as researchers attempt to deal with these challenges. Will artificial general intelligence as broad and deep as human intelligence emerge, or will fundamental limitations become evident that will require further breakthroughs? Although experts continue to debate these questions, what is clear is for the foreseeable future AI technology will continue to advance and applications will continue to emerge that will equal or exceed human levels of performance across a broad range of applications. AI applied to the military domain will lead to improved ISR, command, control, communications, and logistics support. AI will enable new capabilities in autonomous systems and robotics, and will allow the military to develop new platforms and associated concepts of operation. AI “warriors” in electronic warfare, cyber, and grey-zone scenarios, capable of working at scale and speed, will emerge. Indeed, AI has the potential to enable powerful new warfighting systems that synergistically exploit the unique capabilities of humans and machine working together. Moreover, AI applied to science and engineering will accelerate both of these fields, and lead to faster development and deployment of new warfighting capabilities.
After a brief introduction to AI, this lecture will present a generic AI-based system and its basic components, and will give some examples. The role that human-machine teaming plays will be emphasized. The talk will then describe some of the pressing challenges in applying AI to national security applications. To illustrate progress in addressing these challenges, we will present state-of-the-art AI application examples drawn from work currently being conducted at MIT Lincoln Laboratory. The talk will conclude with a discussion on strategic directions to maintain leadership in the application of AI to national security.
About this Lecture
The Massachusetts Institute of Technology Lincoln Laboratory (MIT/LL) "Defense Technology" lecture series will be conducted at the U.S. Naval War College by experts from the world-renowned MIT/LL.
MIT/LL research scientists will provide five LOOs in September and October covering these topics: Cyber Security Threats and Technologies, Military Space, Artificial Intelligence, Communication Systems, and Autonomous Systems. Please check back for changes to upcoming dates and venue.
Please Note: Students and faculty who attend the entire series will receive an MIT/LL certificate of completion.