“Explainable AI is a game-changer for the future of critical decisions”
French-based aerospace and defense engineering firm Thales has acquired Cincinnati, US-based Psibernetix, creator of the human-crushing aerial combat AI system, ALPHA for an undisclosed sum.
Psibernetix’s ALPHA controls fighter planes in simulated aerial combat training missions. It is powered by an in-house designed, machine-learning algorithmic process called Genetic Fuzzy Tree (GFT).
Dr Nick Ernest, of the University of Cincinnati’s College of Engineering and Applied Science department, founded Psibernetix in 2016: the company was part sponsored by the university and the US Air Force Research Laboratory.
The capabilities of the ALPHA system have been praised by US Colonel Gene Lee; a United States Air Force Air Battle Manager and Adversary Tactics (Aggressor) Instructor.
When the Colonel fought in simulated combat engagements against AI controlled by the US Air Force Research Laboratory’s current technology he quickly defeated all foes. However, when he faced off against ALPHA controlled fighters he was consistently defeated.
“Even after repeated attempts against the more mature version of ALPHA, not only could he not score a kill against it, he was shot out of the air by the reds every time after protracted engagements. He described ALPHA as “the most aggressive, responsive, dynamic and credible AI (he’s) seen-to-date,” a research paper wrote by ALPHA’s creators states.
Training complex AI systems that have to make numerous calculations in quick instances is computationally and financially expensive, yet in the academic paper its creators note that: “The very lightweight nature of the GFT, in combination with Psibernetix’s efficient fuzzy logic module, PsiberLogic, and utilization of the Cython computer language in computationally costly procedures allows Psibernetix to accomplish this task with a budget desktop PC.”
Gil Michielin senior VP of Avionics at Thales commented in a release that: “Having certified, explainable AI is a game-changer for the future of critical decisions in safety driven markets. With this new capability, Thales will create truly unique and differentiated AI technologies that will empower customers to make better, more informed decisions, more quickly.”
Psibernetix Genetic Fuzzy Trees
The system is designed using genetic fuzzy trees (GFT); a subset of genetic fuzzy systems. At their heart, fuzzy systems are methodologies that process linguistic information with built-in parameters and mechanisms that navigate uncertainty while helping the system to evolve.
While most AI systems are numeric-based, GFT’s developed by Psibernetix are language-based and contain rules and if/then scenarios that are able to accommodate an array of variables.
A benefit of basing the system in language is that expertise from experienced designers and pilots can be passed onto the system, such as manoeuvrability and tactical advice.
“The ability to input lessons learned from expert knowledge, such as current doctrines and teachings of expert fighter pilots, combined with the ability to fully optimize these concepts via a learning system is a significant factor in the success of the GFT. This architecture, utilization of fuzzy logic, and Psibernetix’s use of the Python programming language allows the development of these systems to be rapid and very cost-effective,” the researchers state.
This is the first generation of a generational programme that the team at Psibernetix’s expects to improve significantly moving forward.
University of Cincinnati aerospace professor Kelly Cohen stated in a university publication that: “In a lot of ways, it’s no different than when air combat began in W.W. I. At first, there were a whole bunch of pilots. Those who survived to the end of the war were the aces. Only in this case, we’re talking about code.”