"Databricks' $10M Investment in DBRX AI Model Falls Short Against GPT-4"
Originally Published 1 year ago — by TechCrunch
Databricks spent $10 million and two months training its new generative AI model, DBRX, which is optimized for English but capable of translating into multiple languages. However, the model's hardware requirements make it difficult for non-Databricks customers to use, and it falls short of OpenAI's GPT-4 in most areas. DBRX also has limitations in accuracy and multimodal capabilities, and its training data sources and potential biases are not fully disclosed. Despite Databricks' promises to refine DBRX, it faces tough competition from other generative AI models and may be a tough sell to anyone but current or potential Databricks customers.