The promise of ambient voice interaction is that it will free us of having to carry devices. Of course, voice is just one interface. It really should be termed “ambient computing”.
In the 1990s, I was intrigued by the idea of the network computer. This was a diskless computer that just had inputs and a screen but could access office applications. The computer itself was just a networked device with a graphics card and keyboard/mouse peripherals.
Today, this idea seems ancient but was really cool. Imagine all the computing you could access. The drop in PC prices, however, killed the idea. However, we don’t need a desktop, real or virtual, to access different services and these are usually exposed as web APIs. Oh, and everything runs in a browser.
Last week, I needed a Windows machine to run some software and within five minutes had access to a VM using Amazon Workstations.
The same revolution is ahead of us with devices and ambient voice interaction. As the incremental improvements between each new smartphone release gets smaller (think iPhone 3 vs iPhone 4 and iPhone 6 vs iPhone 7 or S3 vs S4 and S7 vs S8), we’ll eventually get to a plateau where companies will need to break out of the model and look at new interfaces. Ambient voice and computing provides that break out opportunity.
The key to unlocking this is user authentication, working out privacy issues, and accuracy of inputs. For the latter, this means refining STT for far field and getting around input errors. Another challenge will be creating non distracting notifications and displays from these systems — whether it’s alerts or projections.
The benefit is that our interaction with technology will be much more natural and tuned to our evolution.