"A kit of hardware and software can be installed into different ground platforms to increase the level of autonomy," an Army engineer told the news outlet for a story published July 25. While the exact program remains somewhat ambiguous, Warrior Maven noted that the latest development follows the Army's effort to produce "leader-follower" algorithms that inform unmanned tactical vehicles to follow manned vehicles around a battlefield.
On several occasions, the Army has said that it will focus on joint manned and unmanned vehicles conducting operations. "The upside is that the autonomous systems reduce combat risks and casualties for troops, and help to manage the many ‘fly by wire' combat environments where humans can't operate all systems effectively," Paul Wallis, Digital Journal op-ed editor at large, told Sputnik News.
Indeed, the Army's technologies "are focused on how you actually navigate and detect your position in a GPS-denied environment where there is challenging terrain or an enemy is jamming you," US Army engineer Osie David said in an interview with Warrior Maven.
Some details of the new software kits hint at the possibility that the "artificial intelligence" kit is not actually a new development, but rather an advanced sensor that collects and integrates intelligence from various inputs before sending it to a user to enable quicker decision making. Sensors have been a focus of US military planners for years. For instance, defense sources have told Sputnik News that the F-35 is essentially a "flying computer" that collects a bunch of electronic data which can be relayed to other units within the joint force. Picture an F-35 collecting image data of an ascending ballistic missile, forming a picture of that projectile in real-time and sending this data to US naval forces capable of launching counter-ballistic missile interceptors.
"Instead of sending bits of information back up to a command post, the autonomy kits can enable sensors to perform detection and object identification in real time… and then push that information up to a human" David said.
According to Wallis, "We need to keep in view the fact that all these systems are untried and will be in direct conflict with a whole new class of weapons, maybe millions of different AI systems. In contact with humans, AI systems may deliver some nasty surprises, injuries and a black market in ways of attacking people, lethal or otherwise."
"There needs to be effective oversight, and much more importantly, proper countermeasures to shut down and control any AI systems in contact with humans," Wallis told Sputnik.