The Defense Advanced Research Projects Agency (DARPA) announces the launch of an AI chemistry program, Accelerated Molecular Discovery (AMD), to develop new molecules for US military capabilities.
“We will rapidly be able to explore the unknown molecular universe to design and produce completely new molecules with specific desired functions.”
On Thursday DARPA announced the AMD program, which aims to develop new, AI-based systematic approaches that increase the pace of discovery and optimization of high-performance molecules. AMD looks to develop safe chemical warfare agent simulants and medicines to counter emerging threats.
While DARPA announced that this program would be used to “counter emerging threats,” it did not specify whether or not the technology would be weaponized.
The AI chemistry program will work directly with the US Department of Defense (DoD) to discover and produce new molecules “for a range of military capabilities.”
“The ultimate goal of AMD is to speed the time to design, validate, and optimize new molecules with defined properties from several years to a few months, or even several weeks,” said Anne Fischer, program manager in DARPA’s Defense Sciences Office.
“We aim to develop the AI tools, models and experimental systems to enable autonomous design of molecules to quickly meet DoD needs,” she added.
A Proposers Day webinar describing the goals of the program is scheduled for October 18, 2018.
The problem that DARPA is trying to solve revolves around current approaches to develop molecules for specific applications, which, according to DARPA, are intuition-driven, mired in slow iterative design and test cycles, and ultimately limited by the specific molecular expertise of the chemist who has to test each candidate molecule by hand.
DARPA Looks to AI for Biological Technologies
The defense research organization seeks to develop new, systematic approaches that increase the pace of discovery and optimization of high-performance molecules through development of closed-loop systems that exploit, build, and integrate tools for:
- Extracting existing data from databases and text.
- Executing autonomous experimental measurement and optimization
- Incorporating computational approaches to develop physics-based representations and predictive tools.
“There are currently around 140 million known molecules, but the space of potential molecules is vastly larger than that, upwards of 1060,” Fischer said.
“In DARPA’s Make-It program, we’ve developed advanced ways to efficiently synthesize known molecules. AMD is taking the next step, where we will rapidly be able to explore the unknown molecular universe to design and produce completely new molecules with specific desired functions.”
The AI chemistry program calls for teams of researchers to develop AI-based, closed-loop systems that automatically extract existing chemistry data from databases and text, perform autonomous experimental measurement and optimization, and use computational approaches to develop physics-based representations and predictive tools.
Government partners will evaluate performer developments and test their ability to identify new molecules with specific combinations of functional properties that may be relevant to specific DoD application requirements.
DARPA Views AI as a Collaborative Partner
DARPA is building an Artificial Intelligence Exploration (AIE) program to turn machines into “collaborative partners” for US national defense.
Machines may serve as trusted and collaborative partners
The Defense Advanced Research Projects Agency (DARPA) is looking to make artificial intelligence more trustworthy through a series of projects that test an AI’s capacity for self-evaluation and that make the AI “show its work” for human evaluation.
DARPA’s fundamental and applied R&D AI programs are aimed at shaping a future for AI technology where “machines may serve as trusted and collaborative partners in solving problems of importance to national security.”
Announced on July 20, the AIE program will enable DARPA “to fund pioneering AI research to discover new areas where R&D programs awarded through this new approach may be able to advance the state of the art. AIE will enable DARPA to go from idea inception to exploration in fewer than 90 days.”
The AI Exploration program is one key element of DARPA’s broader AI investment strategy that “will help ensure the U.S. maintains a technological advantage in this critical area,” according to the DARPA announcement.