May 2024 Summaries

AI21 Labs, a company that builds cutting-edge language models (LLMs) for enterprise clients, recently scored the highest transparency score among commercial general purpose LLMs on Stanford University's Foundation Model Transparency Index (FMTI). The company demonstrated significant improvement in its score by sharing more information about its models and performing well on indicators around customer feedback. This high score distinguishes AI21 from its competitors and underscores the reliability of its generative AI models.

The text discusses a multi-expert problem involving green energy companies, where data is fetched from Wiki API, calendar, and database. It then computes the largest increase in share prices using a calculator and formats the answer with a language model. The process involves challenges such as training discrete experts, interfacing them with neural networks, routing among modules, etc. The text also mentions advantages of Jurassic-X, including reading and updating databases in free language, enabling joining multiple databases, and updating databases using natural language commands.

AI21's Generative AI models, Jamba-Instruct and Contextual Answers, are designed to automate and streamline the term sheet creation process for financial institutions (FIs). The solution leverages a two-step process: information extraction using AI21's Contextual Answers model and term sheet generation with Jamba-Instruct Foundation Model. This approach not only saves time but also enhances the reliability and accuracy of financial agreements, making it essential for success in the complex world of finance.