It’s been called a “magic wand,” a “wifi scanner with voice recognition,” and “the little gadget that goes grocery shopping for you.” But the Amazon Dash is best thought of as a part of a generation of Intelligent Assistants that we, members of humankind, should start getting ready to master.

As background, the people who pay the world’s largest retailer the $299 annual fee to be a member of AmazonFresh and who have the good fortune of living in selected neighborhoods in Seattle, San Francisco or Los Angeles can receive a new order entry device for free. The device, as depicted in the video below, is about the size of a TV remote. It has an embedded barcode reader (using LEDs instead of lasers, according to the company) and it has a microphone that supports enough speech recognition capabilities to enable users to dictate their shopping list by entering orders in their own words. “Guitar strings” and “strawberry yogurt” and “apples” are the examples in the video. They are pretty easy to disambiguate.

Have a look:

The device hooks up to a home’s WiFi network. Orders are completed through the Amazon.com Web site  on a PC or the Amazon app on a mobile device. It is unclear from the video whether Amazon uses Dash as anything more than an order entry device, but I can definitely see how a little dialogue management, natural language processing and artificial intelligence will transform it into a full-fledged Intelligent Assistant. Amazon has already received a process patent for applying “predictive analytics” to support “anticipatory shopping,” so it is anybody’s guess how long it will take for Dash to start making suggestions like “aren’t you forgetting the… ‘strawberry yogurt’ or ‘apples’?”

This seems like a natural evolution of Amazon’s recommendation engine. The seeds were planted more than a decade ago for the Web site, when it was just recommending books or dry goods. It is interesting to see how Amazon is embedding speech recognition into TV remotes/set-top boxes and e-commerce Web sites. Based on reviews, there is a definite “wow” factor but there are still some kinks and inconsistency that early adopters are encountering. In this article in The Verge, Dan Siefert points out that the “voice search” is “fast, easy and accurate, but limited to Amazon’s own content.” That means that you can’t use the remote’s voice control once you get into a video game or other streamed application.

Integrating fast, accurate, natural language speech recognition into the user interface of the appliances and devices that a person uses every day is a major step forward. Dash has a limited geographic footprint at this time, but it represents a great leap forward for intelligent assistance. In the guts of Amazon are Yap and Ivona to handle speech rec and text-to-speech rendering and Evi (TrueKnowlege) to support natural language interactions. It’s just a matter of time.

Categories: Conversational Intelligence, Intelligent Assistants, Articles, Mobile + Location

Tags: Speech recognition, Artificial Intelligence, Amazon, Natural language processing, intelligent Assistants