Editor’s Note: We have been receiving a number of articles lately about AI; it is no wonder, given all the interest and activity AI has been generating. Ralf Llansas focuses here on how AI is transforming the capabilities of mobile phones. Given some of what he describes, we can look forward at the same time to improved capabilities to understand consumer needs and behaviors.
Mobile AI will help automate some of your 35,000 decisions per day
The average individual looks at their phones for more than 2.5 hours and makes about 35,000 decisions, in a single day. The fact is, we often use one to help with the other. Our smartphones help us with the choices we need to make. Even now, Artificial Intelligence (AI) is an enormous part of how you use your smartphone. And the support you receive from your mobile phone is about to increase dramatically. AI is making your phone into the world’s best wingman.
Artificial Intelligence is a way of helping computers to ‘learn’ by example from large datasets. AI lets computers acquire information and rules in the sort of way a human does, without being programmed with particular ‘rules’ for every possible eventuality. In essence, AI helps computers to generalize about what might happen next, based on patterns they’ve seen in similar circumstances, before.
What does AI in Mobile Technology mean for marketers?
AI is the marketing ‘Holy Grail’. Every marketing manager’s purpose is to identify customer needs and deliver a product (even if the development of a new one is required) that fits those needs. Mobile technology, especially mobile phones, offer previously never available insights into people’s true behaviors and reveals truths that have previously never been available. AI automates the sifting of that data for insight. The result is a clearer understanding of customer needs and a vivid target for marketers to deliver against.
Just take a look at how the two concepts (AI and Mobile) are being connected in reality. Huawei’s Mate range, Samsung’s Galaxy, Google’s Pixel phones and Apple’s iPhones all now contain special hardware, designed to conduct AI based tasks more efficiently. The capability was in only 3% of phones just last year – but will be part of up to 35% of all handsets sold, by as soon as 2020. (See chart.) Every indication is that we will come to think of the term AI as synonymous with phones, just as other terms did such as a prepaid plan or mobile data.
AI is already a huge part of your phone experience, whether you realize it or not
The number of smartphones shipping with dedicated AI processors is forecast to fast, as the data from one of the world’s largest chip manufacturers, wualcom, shows
Why invest so much? The ultimate goal, of each of these manufacturers, is to improve the experience you as a user, have on their device. The stakes couldn’t be higher. We’ve reached peak smartphone (the point at which the world market is saturated and everyone who wants a smartphone has one). Handset manufacturers are tying together the best possible microprocessor hardware and AI software to deliver the best experience they can and lure customers away from their competitors.
The average smartphone has around a dozen sensors – everything from accelerometers, to GPS, to a microphone, cameras and so on. For years phones have been gathering data on us through the many sensors on a phone. Now AI / ML is learning to use the data they produce.
A host of AI tasks are undertaken constantly in phones, already.
- AI in the camera: AI is already in the camera of most high-end smartphones, more than anywhere else. AI algorithms help identify whether you’re snapping a panorama or a person – and adjusting the type of filtering used to give you the best results. It’s likely AI also helps you by finding the right lens for the light conditions. AI is also behind the facial recognition you might use to get into your iPhone which, again, uses images from the onboard cameras. (In fact – ‘True Depth’ creates a 3D image of your face authenticates for Apple Pay and other payment services.)
- AI in ‘Siri’ and other voice assistants: Today’s smartphones have basic Virtual Assistant capabilities which are improving rapidly. Some speech recognition now offers better levels of comprehension than a human listener, even in noisy environments. Virtual assistants are likely to become a much larger part of our interface with our phone, over time and the natural human language you provide is interpreted by AI.
- Used to allow more human searches of images: Many phones already automatically improve images (for example, removing red eye where it is found) and assist in sorting images you’ve taken and stored in your gallery in a human way – according to friend’s names, for example. Some gallery software automatically pieces together a story of how your weekend went in the form of pictures taken on the day, animations and music.
- AI key contributor to Augmented Reality experiences: The targeted AI processors (often called Neural Processing Engines or something similar) that are found in these dedicated smartphone chips are also used by on-device Augmented Reality experiences, for example, Apple’s Animoji.
- Day to day operations: More fundamentally, AI is behind Google’s core search engine, every time you search from your phone and is being adapted to work behind the scenes in applications as battery life management and security.
Smartphones now have specific microprocessors dedicated to conducting AI tasks
The contribution AI already makes to good on-device experiences, and strategic importance of the field to staying current in a market which has reached peak smartphone has seen each mobile manufacturer accelerate their investments into the field of AI-based user experiences.
This has led to a number of specially designed microprocessors which can conduct the sort of math involved in AI calculations faster and more efficiently – the two most critical requirements for mobile, use less power
Huawei’s claimed to be the first to insert a Neural Engine into their smartphones last year, which they called their Kirin processor / 970 chipset with ‘built-in AI.’ At the time, they claimed that their software could process up to 2000 images a minute. The processor was included in their mate range and will stay as part of their upcoming Mate and Mate 10 product releases.
By splitting tasks across multiple processors, some designed for speed, some designed specifically to conduct AI related tasks, modern smartphones improve performance and battery life.
Significantly, Apple, who has also designed one of these AI targeted microprocessors, are opening up their A12 ‘Bionic’ AI chip – which is also designed to conduct AI tasks more efficiently. Apple included their AI chip for the first time last year, in the iPhone 8, 8X and 8 Plus.
AI is popping up, increasingly, in smartphone software, too
AI is already a key part of Google’s apps on this year’s Pixel 3: ‘Playing now’, for example, is Google’s always on music recognition. The Pixel 3 also has a ‘Call screener’ feature which answers your calls verbally on your beheld, interacts verbally with callers on the other end, summarizes what they say in text format so you can see if you want to call them back. Photobooth’ takes selfies but only when you are your friends are pulling stupid faces.
And Apple is not to be outdone. Access to the chip is also being offered to Apple’s 3rd party app development community. Third party apps will, for the first time, able to use the dedicated processors they want, to conduct their AI tasks, a sure sign of just how serious Apple is about doubling down on the feature.
The mobile phone market is being rebuilt around the concept of AI
It’s no exaggeration to say that the smartphone industry is being revolutionized by and built around the promise of AI. The big opportunity, for now, in AI is to help in guessing what it is we want next. Often, when AI is at it’s best, we users may well not realize that it’s working, things just seem to run a bit smoother. The user interface is of such fundamental importance to the success of phones that each of the major handset manufacturers is investing heavily to ensure they offer the best experience they can and using AI to drive the field forward.
We will come to love our smartphones even more when they eliminate some of those 35,000 decisions we make each day.
And this is only the beginning, as these mobile-appropriate microprocessor chips doing the AI calculations will become cheaper over time and they’re likely to find their way not just to every phone, but every connected device in the world through a 5G connected Internet of Things. (IoT)