Grock 1.5, the latest update from Elon Musk's Grock, brings significant upgrades including improved reasoning and problem-solving capabilities, scoring well on math benchmarks and human eval benchmarks. The context window has expanded to 128K tokens. Grock 2 is in training and expected to outperform current AI metrics. However, the most buzzed-about LLM last week was DBRX from Databricks, surpassing GPT 3.5 and competitive with Gemini 1.0 Pro, especially excelling in code modeling. DBRX is a mixture of experts, with 132 billion parameters at half the cost of LLM270B, offering speed and customization for organizations. The open release of DBRX generated significant excitement in the AI community.