Loud Steps and Quiet Steps
Earlier today (2023.09.20), with all the pomp a big tech firm can muster, Amazon announced that an-all new Alexa voice assistant powered by its own large language model. And, by 200-plus smart home APIs.
Bravo. Smart observers all over the conversational AI industry – @Brandon Kaplan, @James Poulter, and @Bradley Metrock, we’re looking at you – tell us that it’s a long-desired, significant step ahead for the U.S. installed base of nearly 50 million devices.
They’re not wrong. This and other assistant-LLM marriages will give new use, new value, new life to what has been a stale concept, especially for smart home users. Bravo.
Earlier today, with very little pomp, the non-profit Open Voice Network demonstrated operative elements of what will be a universal API for the interoperability of conversational assistants and language models.
This open, universal API has the working title of “Message Envelope,” as it will, when finished, contain specifications that enable the passing of a dialogue (along with history and context) from one assistant to another, along with specifications for authentication, identification, dialog privacy, and security.
In today’s webinar, Open Voice Network speech scientists passed a single dialog back and forth between three independent conversational assistants, each built with a different programming language.
Hmmm.
A loud step for Amazon. A quiet step for the Open Voice Network. Both important steps toward the natural language exploration of the digital world.
On one side, an exciting development in the smart home space. On the other side, exciting developments that, in time. will break down the proprietary walls of personal assistants and provide a secure pathway to the public for enterprise IVRs. And provide the standards-based foundation for conversational AI that works like the web.
On one side, 200+ APIs within the walled garden of a proprietary platform.
On the other side, one open, universal API – defined and designed with recognition that the needs of users and enterprises go well beyond a single platform or LLM. That one size will never fit all.
One loud step. One quiet step. Both big steps toward the future of conversational AI.
To read more about the interoperability work of the Open Voice Network, click here.