This post explores some of the most exciting research directions in AI, such as autoresearch, neural computers, AI safety, or mechanistic interpretability and how philosophy can offer fresh perspectives.
Tutorial on how to rent an external NVIDIA GPU to run experiments with open-source models
Ameliorating strategies for ARC-AGI with multimodal reasoning
This text explores adding sketching to Multimodal Chain-of-Thought (MCoT) reasoning to enhance AI capabilities
Identifying a 3100 BC Mesopotamian contract as an early instance of textuality
This text explores the assignment of gender to humanoids
Tutorial with full code
Tutorial with full code
Introduction to AI ethics that’s less than 1000 words long
Unpublished paper exploring how to use LLMs for empirical studies of syntax
