DARPA seeks to improve AI standards with $1m Ditto project
The US Defense Advanced Research Projects Agency (DARPA) has issued a solicitation for industry partners in its Ditto project, aimed at improving the abilities of machine learning and artificial intelligence.
The agency’s researchers are hoping to make machine learning faster and more accurate by introducing hierarchical ‘thinking’ into AI and machine learning to overcome current faults in AI/ML architecture, which bypass real-world functionality and are typically trained in isolation.
DARPA researchers said, “Modern machine learning algorithms have proven to be excellent mathematical stand-ins for real-world functions, yet suffer from two key drawbacks: lack of meta-cognition, and lack of composability.”
The Ditto: Intelligent Auto-Generation and Composition of Surrogate Models project (Ditto, for short) plans to develop an automated software framework that can recognise microelectronics system design and train machine learning tools that take account of subsystem components, organising them hierarchically, to allow engineers to spot faults in AI earlier and make decisions earlier. The core purpose is to mitigate risk in critical military applications.
A statement from DARPA said: “Today's system modeling technology can be slow, cumbersome, and not always accurate. The DARPA Ditto program seeks to develop an AI framework that can learn to generate surrogate models for different components of a complex system intelligently, aggregate these models while maintaining and communicating surrogate accuracy and coverage, and then integrate these models into one design.”
In inviting industry proposals, DARPA outlined three system design categories: integrated circuits (ICs), mixed signal circuit boards or networked distributed systems. The agency aims to enable “rapid full-scale simulation” by applying a range of third-wave AI techniques to its framework.
It said: “Then the project will develop a proof-of-concept framework with meaningful performance gains in a full-system simulation. The entire Ditto project should be worth about $1million.”
Companies gave until 6 November 2020 to submit proposals.