We’re all spending extra time chatting with our computer systems. This can be a converting interplay with machines. Amazon’s meteoric climb up the Apple App Retailer charts will have to be the entire evidence you want that the age of voice is right here.
Hello Siri, talk to Alexa
Whilst lots of the tech press concerned itself in writing unending Amazon Top day purchasing lists, Brian Roemmele was once looking at what was once taking place in tech.
He discovered that a lot of the techniques offered on Amazon over the previous couple of days have been spoken-word techniques that paintings with Amazon’s Alexa app.
To get a way of what number of offered, he watched as App Retailer downloads of the Alexa app climbed the charts.
Simply how a long way the app complicated is other in several countries, however the place I’m on an iPad the App Retailer presentations me the Alexa app now sits at Quantity 16.
To position this into viewpoint, that’s two puts at the back of the preferred however reviled Fb app and one forward of the ever-so-widely-used Google Maps app.
You don’t want to be a sensible assistant to determine what’s taking place right here.
No longer handiest does this counsel that tens of 1000’s of shoppers bought Alexa-enabled merchandise throughout the previous couple of days, however it additionally presentations that spoken interactions with sensible machines are turning into normalized.
We’re all doing it. We are all the use of Siri, Alexa, Cortana, and everybody else.
Opps. I’m very sorry people for no longer being correct. Within the mins it took me to Tweet this final understand of the Alexa App going to no 1. It moved from quantity 59 to 43!
No app has moved this speedy so far (10 minutes) at the Apple App Retailer.
🗣🚀🏆 https://t.co/3JPpbHOps2 pic.twitter.com/fmfbJlRYfQ
— Brian Roemmele (@BrianRoemmele) July 17, 2019
However, what about scale?
Amazon has a large benefit on the subject of deployment.
The corporate has additionally idea properly round voice.
The creation of Alexa Talents manner the assistant doesn’t essentially want to be outlined as a device for buying groceries or content material get admission to (asking it to play a film or a tune), however can also be prolonged to be used in different domain names.
That’s a bonus at this degree of voice/gadget interplay, a second outlined through brief requests.
Alternatively, as those interfaces scale, this early benefit would possibly expire.
As a result of Amazon doesn’t in reality have a platform (but).
It could actually make one, I assume, however it could take an excessive amount of effort to empower a country of builders to construct a global of apps for tens of 1000’s of utilization circumstances.
It could additionally require the advance of a novel working machine, and the ones are arduous to construct.
Take Photoshop: Enabling Photoshop for voice regulate on a Mac is one thing that has already came about, thanks to Apple’s Voice Control tech.
Voice Control is absolutely Apple’s Royal Flush in the voice/machine age.
Amazon can’t beat that without making a huge investment in technology, development and time, which means Alexa can’t scale.
But it can partner.
What about context?
The thing about spoken-word interfaces is that they are – on the surface – limited.
The commands you can utter are confined to simple decisions, “get this”, “what’s that”, “take a note”, “play a song”, “lights out”…
At least, that’s how it seems.
But this is not the case.
Because the machines are becoming smarter and more capable of contextual analysis.
This helps them predict what you want, understand what you mean, and then deliver a result to fit that context.
AI logarithms such as KNN, LVQ, SVM, and Random Forest mean machine intelligence is becoming more capable of making predictions based on multiple data inputs.
These are the kinds of technologies computers already use to detect spam mail or to figure out the difference between your favorite morning or evening auto-generated playlists.
This developing comprehension around contextual AI means technology inevitably gets better at fulfilling more complex requests by predicting what you need.
Even Siri is getting better at this. It’s getting pretty good at it. Proactive?
Don’t bet against Apple here
Apple has multiple advantages around this.
It’s focus on privacy is extremely attractive to AI developers who now understand the magnitude of what they are building and the dangers of allowing AI to become a gatekeeper to dystopian visions.
Believe it or not, most AI developers just want to make the world a better place rather than take the golden coins while marching humanity to oblivion.
Apple also has the existing platform, millions of developers, a range of essential applications (both its own and those from third parties) and relevant hardware technologies – those A-series processors and their built-in neural chips are passports to mobile AI.
Apple’s processors are fundamental to the future of the company.
And they already smoke the competition.
Put it this way: While Apple’s ’99 Power Mac G4 supported a gigaflop of performance, the iPhone XS some readers carry in their pocket delivers around 5 teraflops.
That’s a huge amount of processing power, more than enough to handle learning vector quantization calculations, as Apple smartphones (I think) already do each time they assemble and recommend a new collection of images in your Photo album.
Memories are precious, but delivering them is AI.
What I’m arguing here is that the capacity to exploit the hardware and software on Apple devices for deployments of AI intelligence mean your devices will be able to indulge you in increasingly complex sets of tasks.
The best way I can explain it is to think about Siri Shortcuts on steroids.
I’m limited in what I can see, but even I can visualize a one-word command to invoke a series of complex Photoshop edits, all on your iPhone. Or a series of words to grab text from one place, automatically lay it out, summon images, and create a newspaper, website or magazine. Or to download data from one or more places and put it through various forms of visual analysis.
Or to do whatever it is you already do in work or play.
All using a word or two.
Words that the AI automatically contextualizes using multiple criteria to your needs, removing complexity through accelerated symbiotic understanding.
Hopefully you see where I’m going.
Alexa also has an advantage
In terms of deployment and familiarity, Alexa is clearly winning consumer wallets.
Despite which, I seem to recall seeing data claiming Siri is now the most widely used voice assistant in the world.
The problem for Alexa is one of scale – though partnership may also be an opportunity.
If there’s a sticking point for deeper partnerships between Apple and Amazon over voice, it’s going to be around user privacy.
It is, after all, not impossible to figure out how to create Siri Shortcuts to initiate actions in Alexa or Google (convoluted as that may seem), but user privacy can’t be eroded as a result.
It’s possible the two companies can find some way to resolve that and other business challenges they have found.
Perhaps all they need to do is respect each other’s domains – for Apple, its about platforms, while for Amazon, it’s about retail sales, I suppose.
Because voice interfaces are not going to disappear.
Nor will they forever be defined by what they can do today – they will become increasingly contextualized and voice will be one of multiple ways in which they can be invoked.
But they’re going to need platforms to thrive.
They’re going to need developers prepared to deliver solutions to real-world tasks beyond the mundanity of shopping.
Alexa’s rapid climb up the App Store charts shows this may be Amazon’s game to lose – but really this shows that voice interface game of thrones has now truly begun.
What do you think, Siri?
Please follow me on Twitter, or sign up for me within the AppleHolic’s bar & grill and Apple Discussions teams on MeWe.