The field of artificial general intelligence (AGI) is witnessing significant developments, with a focus on creating systems that can adapt to various tasks and domains. Researchers are exploring innovative approaches to achieve generalistic scientific reasoning, such as hierarchical problem-solving and neuro-symbolic methods. These advancements have led to the creation of systems that can perform at expert levels across multiple disciplines, including mathematics, physics, and chemistry. Furthermore, there is a growing interest in developing systems that can convert natural language into formal representations, enabling more efficient and explainable problem-solving. Noteworthy papers include: CellARC, which introduces a synthetic benchmark for abstraction and reasoning, and SciAgent, a unified multi-agent system for generalistic scientific reasoning. Additionally, Vector Symbolic Algebras for the Abstraction and Reasoning Corpus presents a cognitively plausible solver for the ARC-AGI benchmark, while Lumine introduces an open recipe for building generalist agents in 3D open worlds.