Rocket Scientists Hooked Up ChatGPT to the Controls of a Spaceship, and the Results Were Not What You Might Expect
Briefly

In a unique research study, scientists from MIT and the Universidad Politécnica de Madrid discovered that OpenAI's ChatGPT performed unexpectedly well in a simulated spacecraft navigation task. The Kerbal Space Program Differential Game Challenge involved various autonomous agents maneuvering spacecraft in competitive scenarios. ChatGPT achieved second place in the challenge, surpassing the expectations of the researchers. The results indicated that with minimal prompts, ChatGPT could orient and maneuver a spacecraft effectively. This research was conducted prior to the release of GPT-4, hinting at potential advancements in AI capabilities for space operations.
Researchers from MIT and the Universidad Politécnica de Madrid found that OpenAI's ChatGPT excelled in a simulated spacecraft navigation challenge, achieving second place.
The Kerbal Space Program Differential Game Challenge tested various AI agents' capabilities in maneuvering spacecraft, revealing ChatGPT's unexpected competence when operating autonomously.
Read at Futurism
[
|
]