Possible Worlds Research

 Possible Worlds Research is a non-profit initiative dedicated to researching and developing tiny AI models, with a focus on Natural Language Processing. Our community is made of researchers and developers who are critical of dominant approaches to machine learning. We think that current AI systems – in particular Large Language Models – create more problems than they solve, ranging from the propagation of harmful biases to exploding energy consumption. Our mission is to build and share intelligent algorithms in a responsible, caring and sustainable way. To do this, we will maintain a dedicated GitHub organisation offering ‘tiny AI’ modules under open-source licenses. We will also foster community engagement around issues of ethics, climate protection and access to technology.