AI Inspired by HAL 9000 Managed Not to Kill Its Simulated Astronaut Crew

The Stanley Kubrick film 2001: A Space Odyssey is so iconic that you probably know some of the major plot points even if you haven’t sat through its 2 hour and 41 minute runtime. The watchful red eye of the artificial intelligence in the film, HAL 9000, is the symbol of modern fears that AI will someday realize that humans are the problem and will kill us all. AI researcher Pete Bonasso and his colleagues ignored that bad omen and used the murderous program as inspiration for their own software called CASE (Cognitive Architecture for Space Agents). In a newly published study, the researchers showed that unlike HAL 9000, CASE successfully kept its crew alive during simulations.

CASE operates on three layers, each with its own list of designated tasks from controlling robotic hands and eyes, to perform routine tasks and checks, to planning and rescheduling based on changing priorities in a planetary base simulation. The layers connect to an ontology server which acts as the reasoning center, making sense of the data and putting all of the parts in context. A key difference between CASE and HAL 9000 is that the new software is programmed to share plans with humans and obtain consent. “Our colleagues and NASA counterparts are not concerned that our HAL might get out of control,” Bonasso told Space.com. “That’s because it can’t do anything it’s not programmed to do.” 

The longest simulation that CASE was able to run was 4 hours. The researchers say much more testing is needed before the AI would be able to run a real base with actual astronauts. “Though CASE is impressive,” said Bonasso, “it’s not the fully realized HAL from ‘2001: A Space Odyssey,’ nor is it Lt. Commander Data from ‘Star Trek: The Next Generation…its capability is very narrow, focused on events occurring on a planetary base. While it can keep the life support systems running, it has no idea who won the last presidential election.”

That last part is probably for the best. The last thing we need is an AI sharing its political views on Twitter.

Sharing is caring!