The editor of Downcodes will take you to learn about the latest artificial intelligence in Minecraft - AIRIS (Autonomous Intelligent Enhanced Inference Symbols). It is not pre-programmed, but masters game skills through continuous learning and practice in the game, and explores the world of Minecraft from scratch. This marks a major breakthrough in the autonomous learning ability of artificial intelligence in complex virtual environments. Its learning process and application prospects are worthy of our attention.
Early versions of AIRIS were tested in a simple 2D grid world puzzle game environment. However, in order to test the system, developers need to test the system in a more complex and open 3D environment. Minecraft fits that description perfectly, it's a very popular game and has all the technical requirements to plug AI into it.
AIRIS works by taking two types of input from the environment and a list of actions it can perform. The first type of input is a 5x5x53D grid surrounding the agent's block names. This is how an agent "sees" the world. The second type of input is the agent's current coordinates in the world. This gives us the option to specify where we want the agent to go.
AIRIS starts in "Free Roam" mode and attempts to explore the world around it. Build an internal map of where it has been, which can be viewed using the included visualization tools. It learns how to navigate the world, and when it encounters obstacles such as trees, mountains, caves, etc., it learns and adapts to them.
Successful use cases for AIRIS may include automated error and stress testing of software. It's assumed that AIRIS can run throughout Fallout 4, creating bug reports when interacting with NPCs or enemies, for example. While quality assurance testers will still need to check what the AI records, it will speed up a tedious and frustrating process during development.
The emergence of AIRIS is the first step towards a virtual world where artificial intelligence can learn autonomously in a complex, omnidirectional world. This should be exciting for AI enthusiasts as a whole.
The successful application of AIRIS is not limited to the gaming field. It has huge potential in software testing, automation, etc., indicating the application prospects of artificial intelligence technology in a wider range of fields. The editor of Downcodes looks forward to the future development of AIRIS and continues to pay attention to the latest developments in the field of artificial intelligence.