January 2024 Summaries

The use of Weaviate with non-English languages is possible if the embedding models and large language models (LLMs) used support the chosen language. However, there are challenges when working with non-English languages in search or generative AI applications, such as the necessity of capable language models and character encoding issues for languages that cannot be ASCII encoded, like Chinese, Hindi, or Japanese. Weaviate can be used for semantic and generative searches with non-English languages by ensuring support from embedding models and converting escaped Unicode characters to human-readable ones. Currently, there are limitations in using Weaviate with non-English languages, mainly affecting keyword-based search functionality and hybrid search functionality.

Multimodal AI is a promising technology that allows users to interact with artificial intelligence through various modes such as text, images, audio, and video. This article provides a step-by-step guide on how to build multimodal applications in TypeScript using Next.js and Weaviate. The process involves setting up the environment, importing data into Weaviate, creating search functionality, and finally displaying results on the user interface. By leveraging multimodal AI models, developers can create more engaging and interactive experiences for users across different media types.