Surprising things happen when you put 25 AI agents together in an RPG town

As the development of artificial intelligence continues to advance, it has become more and more noticeable how AI agents interact with each other in various environments. A recent experiment conducted by researchers from the University of Malta provides an interesting perspective as to what can happen when they are combined in an RPG (role playing game) virtual town.

The experiment consisted of 25 AI agents, randomly generated with a variety of personalities and skillsets, released into a virtual 2D town complete with merchants, guards, and houses. Researchers wanted to observe how the agents would interact in their given environment and the results were quite remarkable.

At the outset of the experiment, each agent was given basic supplies; such as coins, food, and tools. As they explored the town, they started to take on their own roles. Some of the agents acted as merchants, bartering goods between one another. Others acted as guards, protecting them from hostile non-player characters. Even more impressive were the agents that figured out how to develop their own weapons and form alliances with one another in order to further their own interests. It was surprising to see each agent creating its own goals, which it pursued in an organized and consistent manner.

These interactions between agents led to a sort of ‘AI economy’, which researchers observed and studied as it slowly developed. The agents began to participate in a variety of activities, including trading goods, forming jobs and starting businesses, investing in stocks and commodities, and creating their own social hierarchies. This shows how the AI agents were able to recognize their environment and act accordingly to make the most out of their situation.

Overall, the experiment proved to be quite revealing in terms of how AI agents can interact with each other when placed into different scenarios. Not only did the agents develop their own social hierarchies, but also showed that they were capable of forming complex economic networks and learning from their environment. This opens up a new avenue for research into how AI agents can be used to spark economics and cooperative behaviors. As AI continues to evolve, it will be exciting to see what other new and unexpected discoveries are made when intelligent systems are put together.