Amazon seems to be continuing its strategy of expanding Alexa from the ashes of its aborted Fire Phone project – this week expanding its third-party support for adding Alexa to other devices. While the Echo seemed like a surprise hit for the company, it’s still a little baffling that it hasn’t made more of an effort to sell it outside the US.
But a conversational AI-based digital assistant is a very complex feat of software engineering. Running on a central cloud computing installation, the system responds to the voice-inputs recorded on the end-device – processing them and then responding to a question or command in the way it perceives to be correct.
This decision can evolve over time, and take into account the fast-moving nature of web content, so that users can just as easily ask Alexa to turn down the kitchen lights as they can query a recipe or find out the latest score update for their favorite sports team.
Although Amazon will never confess it, Alexa and the Echo are designed to make it easier for Prime customers to spend more money through the Amazon marketplace. They are gateways to the retail giant, and as such, are designed to be as simple to use as possible – with some of the earliest use cases for the Echo involving adding products to a shopping list.
Essentially, Alexa is being positioned as the more open AI platform. Apple retains an iron grip over Siri, which is only available on Apple hardware, with integrations through iOS APIs. Microsoft’s Cortana is in a pretty similar position, but tied to the Windows 10 platform rather than specific products.
Elsewhere, Android’s Google Now system would appear to provide an open alternative, but developers are then tied to the Android platform and the associated Google licensing terms that come with it. The core components of Android are open source (AOSP), but the Google apps themselves that form the core value offering in the smartphone arena (Google Mobile Services) are tied to a bundled license, where the developer can only include one Google app if the rest are also included as standard on the device.
That arrangement may change, but at the moment, it seems that an Android-based Echo-derivative would be quite bloated on the software side of things. AOSP can provide a very good platform on which to build Android devices, especially at the lower-end of the hardware capability spectrum. Whether AOSP plus GMS ends up being so dense a software environment that it negatively impacts the BOM cost of one of these Echo-clones depends on strides in low-cost SoCs, but it’s a consideration that needs to be kept in mind.
In the past week, we have learned that Amazon has expanded its Alexa Voice Services (AVS) developer tools, which were launched towards the end of last year. The update means that developers have access to a number of new features, which include the ability to use Alexa to adjust playback volume, media controls, and set alarms.
The core function of AVS is to allow those third-parties to build hardware that houses the Alexa interface – allowing the device to act as the gateway to Amazon’s cloud and monetization strategy. The update includes “architectural improvements that include updated APIs and messaging structures, Alexa app support, and the ability to send server-initiated messages.”
Amazon has also announced that it has made the $100m Alexa Fund available to businesses looking to incorporate Alexa into their designs. Early investments have included Petnet (automatic feeder for pets, voice queries), Musaic (smart HiFi system for whole-home audio, aiming for voice control), and Mojio (connected car adapter for smartphone telematics).
In addition, AVS has also been ported to run on the Raspberry Pi – a considerably cheaper platform compared to a smartphone chipset, which makes the aforementioned argument against Google Now a little more compelling. Already available on GitHub, the code should allow hobbyists access to Alexa – which could be the basis of the next garage-innovation that takes the consumer electronics market by storm. Alternatively, it’s another useful source of data for Amazon to use to refine Alexa.
At the launch of the preview of AVS, Scout announced that its home security products would use Alexa, and Toymail said that its messaging-enabled children’s toys would also be making use of the new toolkit.
by Peter White, Rethink Research, contributor to AI Trends