Databricks' US$1.3 Billion MosaicML Acquisition: The AI Deal Wave Is Just Beginning

Cloud data giant Databricks, itself a super-unicorn, announced a US$1.3 billion acquisition of MosaicML, an AI startup less than three years old. MosaicML was founded in 2021 and has 62 employees. It has disclosed only one financing round, at a US$220 million valuation. Databricks' acquisition represents a sixfold increase in valuation. Founded mainly by a group of Intel executives, MosaicML focuses on improving enterprise AI efficiency, particularly in generative AI. It provides an AI platform that lets enterprises easily train and deploy models in secure environments while reducing the costs of training large AI models. Databricks plans to integrate its Lakehouse platform with MosaicML's technology. The combined product will let enterprises simply, quickly, and inexpensively train and build generative AI models using their own proprietary data, developing customized models while retaining control and ownership of that data. MosaicML's automated optimization of model training promises training speeds two to seven times those of traditional methods. Near-linear resource scalability means models with billions of parameters can be trained in hours rather than days. According to Databricks, its and MosaicML's platforms and technologies will significantly reduce enterprises' costs of training and using large language models (LLMs), potentially to around a few thousand dollars. Amid the current AI boom, technology giants' acquisitions of generative AI startups have only just begun.