The space agency is working on a generative A.I. system that would let astronauts and mission control talk to vessels and perform experiments, reports the Guardian. The technology is currently slated to deploy on the Lunar Gateway, a space station that will orbit the moon as part of the Artemis lunar program and is scheduled to launch in 2028.

“The idea is to get to a point where we have conversational interactions with space vehicles and they [are] also talking back to us on alerts, interesting findings they see in the solar system and beyond,” said Larissa Suzuki, a visiting researcher at NASA during a meeting on next-generation space communication at the Institute of Electrical and Electronics Engineers (IEEE). “It’s really not like science fiction anymore.”

Let’s hope not. In Stanley Kubrick’s seminal 1968 film 2001: A Space Odyssey, the spaceship’s A.I., named HAL (Heuristically Programmed Algorithmic Computer) suffers a series of malfunctions (hallucinations?) that lead to (spoiler alert) the system killing almost everyone on board.

Suzuki theorized the real-world NASA system could be used to detect, and perhaps fix, data transmission glitches, since sending an engineer up whenever something goes offline is impractical. Additionally, astronauts and mission control will be able to use the interface to quickly obtain relevant information, rather than digging through technical manuals.

One of the challenges, though, is to get machine learning operating in space, without huge data dumps to and from Earth, which could slow operations down. One idea Suzuki, who also works at Google, floated was having a fleet of rovers sharing information locally.

A separate page hosted by NASA says it will use A.I. and machine learnings “to manage and operate engineered systems to facilitate long-duration space missions.”

Just to be safe, though, NASA might want to avoid sending up any astronauts named Dave until it’s sure all the bugs are worked out.