About Us

We are an open source artificial intelligence research collective building stuff for fun.
Our group collecting has over 30 years of experience working with machine learning projects in research and production.
Our main project now is to create the most advanced,realistic, and fun to use artificial intelligent agent simulation software, our current version is at: https://github.com/jtoy/replicantlife

What does advanced and fun mean?

We want our agents to be advanced and smart like a human. Agents should make intelligent decisions based off the information they have and from previous interations. Agents should learn over time just like humans do.

Advanced

Up until LLMs came on the scene, most AI agents have not been very smart, but instead had lots of hand tweaked rules. We want agents to make decisions like a human, which means smart, dynamic, and realistic.

Fun

- We want to have fun building this and we want users of the software to have fun using it as well. TOthis end, we have added many flags to change the behaviour of the agents. You can use the agents in many ways such as chatting, assisting you, role playing, etc. All of this is open source.

Our first paper is out: link
We have several research papers in the works.
If you are looking to work on some cool projects or have any feedback, please contact us at contact AT replicantlife.com