In the last twelve months, there has been a notable push by manufacturers towards the AI capabilities of smartphones for a whole array of different purposes. But what exactly is AI and what effects does it have on the day-to-day use of your smartphone?
AI: What is it?
First and foremost, AI quite literally stands for ‘Artifical Intelligence’ and refers to the ability of computing hardware to exhibit the cognitive functions usually associated with human brains; thinking, learning and problem-solving. In essence, machines are given intelligence that is not too discernible from our own, through a system of information and applications that are built into them. As such, this hardware can essentially think for itself and use the data that it has been fed to come to certain conclusions.
As you can imagine, these systems process absurd amounts of data in very quick times, so are extraordinarily powerful and hugely complex. Most newer flagship devices have hardware that is optimised for AI, running chips that are usually referred to as a neural engine or the neural processing unit. Importantly, these chips have been designed for the fast processing requirements that we discussed previously.
The AI capabilities of smartphones have been utilised for an array of different needs since its introduction. Handsets work to distinguish patterns in behaviour, learn and then determine what you need accordingly.
AI in Smartphones: Where and How?
#1: Camera
One of the most common uses of AI technology and probably where you’ve seen the term most often, is in relation to the camera tech of a smartphone.
Manufacturers in this instance use machine learning to create image recognition engines which have been trained using millions of different images. Following this, the AI technology is then able to automatically identify a number of different scenes, available light and in some instances, different angles of a scene.
As such, smartphone cameras with AI support can recognise and distinguish between a plethora of different environments and adjust the phone’s settings automatically; including lighting, contrast, exposure and saturation. This leads to Instagram-ready photographs that don’t need much after-shot editing or tweaking. Indeed, in many cases, image recognition is so well-refined that a camera can distinguish a cat from a dog and adjust settings accordingly.
AI technology is equally beneficial when in portrait mode, in which the details of a subject are refined, and the background is slightly blurred for that polished and artistic effect. AI object recognition is able to distinguish between the edges of a subject’s head – including various hairstyles – and the background, applying the blur where needed. In this way, AI technology isn’t actually being used to take the photo, but rather to process the data that is captured when the image is taken, improving the overall quality.
On a number of newer flagship models, you are able to use the AI tech of the camera to search for an image in front of you. If you point the camera at a pair of shoes, it will then use that image to make a search online and come up with a list of different shopping results.
Alongside this, AI can also be used to help with low-resolution images, predicting the pixels that are missing for a better full display and zooming in.
#2: Battery
Another way in which AI works is to help to conserve your battery – hoorah!
Again, AI works to monitor your phone usage and adjusts the devices settings accordingly. In this way, the technology dedicates certain power to certain different functions. As such, if you’re not using a particular feature or application, then the phone will not allocate it any unnecessary power. Generally, a lot of battery power is used to update apps that may rarely be used and that aren’t even running, so AI will learn habits and drop apps that aren’t being used.
So too, a common feature for recent flagships is ‘Adaptive Brightness’ which uses AI to learn how you tend to set your brightness settings and replicates it in given scenarios. As such, if you’re in a low light situation, your phone can dim your brightness and subsequently aid towards preserving battery light.
#3: Personalisation
With all of this learning and subsequent problem-solving, the most obvious benefit of AI technology within smartphones is the creation of a smartphone experience that is tailored to the needs and uses of the individual.
Instead of running a generic profile that is created for the needs of many, AI processors instead learn the behaviour of each user and adapt the way in which it works according to owner habits. For example, your smartphone may have learnt what apps you use at a particular time or location, as well as those you use regularly. Have you ever walked into a particular store and your phone flashed up, suggesting that stores app?
#4: Voice Assistant
Most smartphones now come with some kind of voice assistant, whether that’s Siri, Google Assistant or Bixby. You simply ask it a question and it will respond with an appropriate answer and help you to schedule reminders and set alarms. Most voice assistants can also understand context and learn usage patterns to become smarter over time.
Natural Language Processing (NLP) is the terminology used to describe the ability of a machine to understand what is being said to it verbally, understand what has been said to it, work out the appropriate action and then reply using the same language that was used when the question was asked. All of these actions are only achievable through AI.
#5: Face Unlock
Another regular inclusion within recent flagship devices is Face Unlock; allowing you to unlock your smartphone simply by looking at it.
The iPhone X’s True Depth Camera in particular has won a lot of praise and works by taking a number of 3D images of your face, that are subsequently used to authenticate you. This also uses machine learning so that the technology can adapt to certain changes in your face, whether it be weight, hairstyle, expression or accessories to recognise your face more quickly.
6: Other
The capabilities of AI technology are countless and so we could be here all day discussing all the ways in which it might affect your smartphone usage daily.
Google Maps street parking difficulty indicator is a fantastic example, which informs you of how difficult it might be to find a parking space at a specific location. The technology combines crowdsourced data and machine learning algorithms that looks at how long it took for other users to find a space at the same location.
Live translations is also a much-loved use of AI, which analyses a number of different written languages.
When it comes to AI technology within smartphone cameras, then you can’t look much further than the Huawei Mate 20 Pro.
Huawei claim that the Mate 20 Pro camera app can recognise over 1500 scenarios across 25 different categories, adjusting the settings accordingly. Colour saturation within images is arguably unparalleled and the device will even suggest when it thinks you should change camera modes.
For the iPhone Xs Max, Apple created the A12 Bionic chip that was designed specifically by Apple to utilise the processing abilities of AI.
The iPhone X received deserved plaudits for how this was used to take portrait images and within Face ID as we discussed above. However this worked at 600 billion operations per second. In comparison, the XS max is capable of 5 trillion – yikes!
Samsung is another smartphone manufacturer that has become renowned for its photography – and rightly so!
The Galaxy S10 Plus utilises AI to great effect for a number of different purposes when it comes to its camera. Indeed, Samsung has introduced ‘Best Shot’ mode, which is able to look at a scene and suggest where a user should point their sensors for the best image. A plethora of different scenarios can be accurately deciphered and the device will also automatically change to night mode when presented with a low-light scene.