We lead multidisciplinary research to perceive, decide, act, and team well beyond what is possible through human cognition alone, specifically focusing on AI exploration, AI vulnerabilities and defenses, biological intelligence, contextual intelligence, and teaming at the speed of thought.
AI EXPLORATION
Exploring fundamentally new AI concepts, applications, and research vectors
To enable the continued rapid pace of AI innovation, our work explores new concepts for data, learning, and computing, new cognitive paradigms, and new ways of integrating technologies for emerging applications such as AI for scientific discovery and complex design.
AI VULNERABILITIES AND DEFENSES
Enabling safe and assured AI for adversarial and sensitive contexts
For every new national security application of AI that comes online, critical vulnerabilities closely follow. Our research focuses on the structure and dynamics of intelligent systems to identify, detect, and mitigate critical vulnerabilities, adversarial and otherwise, within the much broader AI safety landscape.
BIOLOGICAL INTELLIGENCE
Researching neurally and cognitively inspired algorithms and architectures
The mind and brain have been the state of the art in intelligence for over 500 million years. We perform targeted research at the intersection of neuroscience, cognitive science, and computer science in pursuit of more capable and resource-efficient artificial intelligence.
CONTEXTUAL INTELLIGENCE
Advancing dynamic, multimodal reasoning for real-world decisions
Realizing the potential of AI for mission impact requires intelligent systems that can synthesize knowledge—from text and speech across languages to digital signals across sensing modalities—and reason flexibly about the world.
TEAMING AT THE SPEED OF THOUGHT
Developing human-AI teaming to perceive, decide, and act well beyond human cognition alone
As AI continues to advance, human operators will increasingly become the weakest links in the intelligent systems they rely on. To take greatest advantage of the growing capabilities, we are developing neural interfaces and cognitive teaming approaches to interact with technology, each other, and our environment in new ways.
Research Highlights
News Story
Johns Hopkins APL’s Cyclone Aims to Improve Collaborative Human-Machine Decision Making
Read More >>
To improve collaborative decision making between human and machine, researchers at APL created an artificial intelligence agent called Cyclone and trained it to play a cooperative card game through a unique learning process: train Cyclone to play like a human so it can play well with humans.
Read full article >>
Press Release
APL and the Intelligence Community Tackle Malware in the Age of AI
Read More >>
APL scientists are working with the intelligence community to develop fundamentally new methods to inspect artificial intelligence for Trojans — vulnerabilities that deep networks are exposed to during the AI training process.
Read full article >>
Press Release
APL Developing Noninvasive Brain-Computer Interfaces to Control Complex Systems
Read More >>
With a recent award from the Defense Advanced Research Projects Agency (DARPA), APL researchers will develop a brain-machine interface that will enable the control of complex systems at the speed of thought.
Read full article >>
Press Release
Researchers Directly Observe Concepts in Human Brain
Read More >>
When seeing objects in the world, individuals probably are not thinking explicitly about their semantic characteristics: Is it alive? Is it edible? Is it bigger than a bread box? But activation of these kinds of semantic attributes in the human brain is now directly observable, according to recently published findings from Johns Hopkins University, its Applied Physics Laboratory, and its School of Medicine.
Read full article >>
