Inside the Model Factory: Eiso Kant of Poolside AI Discusses Rapid Model Development

Poolside AI co-founder Eiso Kant explains the company's Model Factory, which can take a model from pre-training to release in eight weeks, and their new Laguna S 2.1 model that outperforms much larger models. The interview covers Poolside's open weights approach, focus on code as a path to AGI, and engineering systems for running 10,000-20,000 experiments per month.