The field of quasi-Monte Carlo methods is witnessing significant advancements with the integration of machine learning techniques. Researchers are exploring new ways to generate low-discrepancy point sets and sequences, which are essential for efficient numerical integration and other applications. The use of large language models (LLMs) is becoming increasingly popular in this area, with applications in program synthesis, model merging, and collaborative co-evolution. Notably, LLM-guided evolutionary program synthesis has been shown to automate the discovery of high-quality quasi-Monte Carlo constructions. Furthermore, the development of neural low-discrepancy sequences is providing a promising approach for generating sequences with minimal discrepancy across all prefixes. In the realm of multi-LLM collaboration, researchers are investigating ways to measure the chemistry among collaborating models and recommend optimal model ensembles. This has led to the introduction of frameworks such as LLM Chemistry and Multi-LLM Collaborative Co-evolution (MCCE), which are enabling more effective collaboration and co-evolution of LLMs. Some noteworthy papers in this area include: LLM-Guided Evolutionary Program Synthesis for Quasi-Monte Carlo Design, which demonstrates the potential of LLM-guided evolutionary program synthesis for automating the discovery of high-quality quasi-Monte Carlo constructions. Neural Low-Discrepancy Sequences, which introduces a machine learning-based framework for generating low-discrepancy sequences. MCCE: A Framework for Multi-LLM Collaborative Co-evolution, which presents a hybrid framework that unites a frozen closed-source LLM with a lightweight trainable model for multi-objective discrete optimization problems.