Microsoft’s AI Innovations: Vector Search and Voice Cloning Via Custom Neural Voice
Microsoft, the tech behemoth, is at it again. This time, they’re making waves at their annual Inspire conference with a slew of AI features headed to Azure. The star of the show? Vector Search. Now, don’t let your eyes glaze over just yet. This isn’t your run-of-the-mill search tool. Vector Search is a machine learning marvel that captures the meaning and context of unstructured data, including images and text, to make search faster. Yes, you heard it right. Faster.
Vectorization, the secret sauce behind Vector Search, is a technique that’s gaining traction in the search world. It’s all about converting words or images into vectors, or series of numbers, that encode their meaning. This allows them to be processed mathematically. In simpler terms, it’s like teaching machines to understand that words close together in “vector space” — like “king” and “queen” — are related. This makes it possible to quickly surface them from a database of millions of words.
But wait, there’s more. Microsoft’s flavor of vector search offers “pure” vector search, hybrid retrieval, and “sophisticated” reranking. It can be used in apps and services to generate personalized responses in natural language, deliver product recommendations, and identify data patterns.
Document Generative AI: The Future of Document Processing
Microsoft is also launching the Document Generative AI solution. This integrates Microsoft’s existing AI-powered document processing services, including Azure Form Recognizer, with the Azure OpenAI Service. This solution ingests files for tasks like report summarization, value extraction, knowledge mining, and generating new types of documents.
For example, using the Document Generative AI, a customer could upload invoices, bills, and contracts to allow employees to ask questions about service guarantees and specific line items. The Document Generative AI solution answers questions in text as well as images and tables, providing citations with a link to the source content.
Voice Cloning: A Double-Edged Sword
In a related announcement, Microsoft revealed that OpenAI’s Whisper model, an automatic speech recognition model, will soon come to the Azure OpenAI Service as well as Microsoft’s family of AI speech services. Enterprise customers will be able to use Whisper to transcribe and translate audio content as well as produce batch transcriptions “at scale,” Microsoft says.
Microsoft also announced the wider availability of Custom Neural Voice, which taps AI to closely reproduce an actor’s voice or create an original synthetic voice. However, with the potential misuse of voice cloning technology, Microsoft has put in place controls to help prevent misuse of the service.
So, that’s the latest batch of news from MSFT. Microsoft is pushing the boundaries of AI, but with a keen eye on ethical considerations. It’s a delicate balance, but if anyone can pull it off, it’s Microsoft.
