QuEST EVENT - This week we will read Lila Davachi's recent paper on memory reactivation (conscious/nonconscious, awake/asleep). 15 March 2024
ACT3-header

Autonomy Capability Team

About the Autonomy Capability Team (ACT3)

AFRL’s Autonomy Capability Team (ACT3) is an AI Special Operations organization whose mission is to Operationalize AI at Scale for the Air Force. Commissioned by the AFRL Commander, ACT3 leverages an innovative ‘start-up’ business model as an alternative approach to the traditional AFRL Technical Directorate R&D model by combining the blue sky vision of an academic institution; the flexibility of an AI startup; and the discipline of a production development company. ACT3 integrates the world’s best under one roof. The goal of the ACT3 business model is to define the shortest path to successful transition of solutions using AFRL’s internal expertise and collaborations with the best academic and commercial AI researchers in the world. Successful implementation may mean new technology or new implementation of existing technology.


    

QUEST TALK VIDEOS

Upcoming Quest TALKS

  • QuEST TOPICS
    Click below for upcoming and past QuEST guest speakers, topics and talks.

QuEST TALK SPOTLIGHTS

* ​The views, opinions, and/or findings contained in this presentation are those of the author and should not be interpreted as representing the official views, position, or policies, either expressed or implied, of the United States Government, the Department of Defense, the United States Air Force or the United States Space Force. The intent of this academic seminar is to discuss publicly available science with thought leaders in the field.


    

What is Artificial Intelligence (AI)?

AI refers to a machine’s ability to gather observations, create knowledge and apply that knowledge to accomplish certain tasks. As we approach an era when most of our knowledge will be machine generated, AI will enable people to use this knowledge effectively. When AI examines an environment and creates knowledge from the observations gathered, it analyzes an abundance of useful information in far less time than the human mind. Examples of AI include speech-processing, robotics, the “auto-tag” feature on photos and self-driving cars.

stock photo of artificial intelligence interacting with humans

Why is AI important to the Air Force?

According to the 2019 National Security Strategy, operational AI will unlock infinite possibilities and enable tremendous efficiencies within the U.S. military. However, to reach this goal, a concrete plan must link AI opportunities to Department of Defense (DoD) requirements. Successful AI requires continuous development and integration. With this in mind, the current path forward requires a scalable solution.

digital graphic of jet fighter

Autonomy Capability Team (ACT3)

ACT3 is a group of experts within the Department of the Air Force leading the development of the Air and Space Force Cognitive Engine (ASCE), an AI network/business model with a single IT platform that operationalizes AI within the Air Force and Space Force. ACT3 has assembled the necessary people, algorithms, data, and computing resources to support the successful fielding of AI capabilities.

stock photo of ai in digital environment

Air and Space Force Cognitive Engine (ASCE)

ASCE is a one-stop, all-inclusive, license free, suite of closely integrated software and services that enable users to develop complex AI solutions at any scale on top of any Kubernetes cluster. ASCE supports development on individual laptops as well as easily conducting experimentation on a Kubernetes cluster. This cooperative suite of tools supports users as they research, prototype, continuously test, improve and deploy their AI products.
 
Designed in close collaboration with a large cohort of researchers within ACT3 and beyond, ASCE is designed to help overcome the common challenges in working with data and steep computational requirements. Specifically ASCE provides an easy to use interactive computational environment with full GPU support. It also provides data packaging, discovery, and delivery. The vision for ASCE is to enable AI to be a tool in the hands of every Airman, with the ability to interact with AI in a do-it-yourself fashion that will result in an exponentially innovative landscape.
 
This unique approach of creating systems uses existing AI to create new, world class AI. Requirements that may benefit from ASCE include business processes such as civilian hiring and contract monitoring, predictive maintenance, automated air combat operations, aircraft damage inspection, humanitarian assistance, disaster relief, and additive manufacturing. Intelligence, surveillance and reconnaissance missions can also benefit from ASCE. More generally ASCE significantly lowers the barrier to entry to using a large scale cluster for computational science.

Endorsements

Here’s what our colleagues have to say about ACT3: 

  • Tom Mitchell, Carnegie Mellon University - "…the system architecture and cognitive designs embodied in the USAF ACT3 platform are fundamental to the future of machine learning and with key attention to continuous learning in machine."​
  • Yann LeCun, New York University, Director of AI at Facebook, and Turing Award Laureate - "Deep Learning is an exceptional mechanism addressing many research questions in machine learning; it is a computational tool for many applied problems, but it is not the “tool” for everything…what it is necessary to understand is how many different types of “computational tools” work together…this is exactly what ACT3 is doing…dynamically combining various intelligent software agents to learn and reason about various classes of real-world problems…ACT3 is advancing critical machine learning and AI research."
  • Marvin Minsky and Patrick Winston, MIT - "Serious advances in the science of AI will require an understanding of the integration of 100s of sophisticated learning and reasoning systems…the ACT3 design provides a creditable means to achieve such an integration…its architecture and multi-representations of knowledge and data extend the state-of-the-art in multi-strategy reasoning and learning." Minsky and Winston both concur that the ACT3 work addresses a famous Minsky quote: "You don't understand anything until you learn it more than one way."
  • Eric Horvitz, Director of AI at Microsoft - "…with the compelling advances in Deep Learning, Reinforcement Learning and many other techniques…the work by the ACT3 team serves the AI/Machine Learning community with significant breakthroughs to integrate various computational AI techniques, thereby obtaining new levels of performance in machine learning…and most critically, engineering the AI scaling challenges due to the dynamic architecture of the ACT3 platform."
  • Vladimir Vapnik and Rauf Izmailov, Columbia University - "The ACT3 engineering is addressing one of the critical concerns, that being, applying context during learning in a situation…such work requires multiple types of representations used in distinct types of diverse reasoning/learning…the Vapnik issue is, without context, the future of machine learning will reach an impasse…ACT3 is obtaining a new paradigm for integrated diverse learning machines working in collaboration…ACT3 provides a dynamic means for many different AI systems to work together providing different perspectives of context…this is the machine learning path to take…"

ACT3 LEADERSHIP

Relevant Links