Recently Apple launched the iPhone 7. Although machine learning was mentioned only a few times the possibilities it will bring are huge. Inbuilt Artificial Intelligence will make our portable devices even smarter. The new camera system in iPhone 7 will perform 100 billion operations for each photo, taken by AI silicon.

Examples of competing AI technology in the mobile phone industry are; Face recognition, search, route selection and navigation in maps, News Feed (from Facebook), image manipulation, personal assistants, translation of languages, virtual and augmented reality.

Screen Shot 2016-09-26 at 16.16.05September 20.th Apple rolled out the new iOS 10 software, available for more than 1 billion iPhone users. The Photo app has been upgraded with neural network AI that let you find images of people, places, groups, occasions etc. based on the phone understanding or recognising these patterns.

Apple Photos is performing about 11 billion operations per photo to figure out who is in the picture, what mood they are in, where and when it is taken, recognising scenery, animals and races, or objects in the picture, activities and many other tags that describe a picture. It then uses this information to make links between images so it can present natural groups of images, from a specific trip for instance. Awesome!

This technology isn’t just suitable for crime scene investigators, it’s deemed suitable for the general public too. Imagine what you can learn from the photo’s you’ve taken!

Apple is adding more AI features in iOS 10, which is no surprise. This is a trend that has been ongoing. AI is embedded in the silicon of the new iPhone 7 is an ongoing trend and we can expect to see more in the years to come.

Please contact us at Cloudnames if you have any questions. You can send us a direct email on hello@cloudnames.com or simply leave your contact details below and we’ll get right back to you.

Do You Need a Website?

Do you need a brand new website? Or, perhaps you need to update your existing one? You can opt for our basic web design package starting from 990 GBP, or speak to an expert today and get a customised quote. If you require more information, simply leave your contact details and we’ll be in touch shortly.

 

The phone is the key
Initially AI solutions was all about cloud technology, simply because the computational power of a smartphone was not strong enough. For example, Apple’s Siri personal assistant is a cloud based service. And, Siri was first launched as an app for the iPhone, then bought by Apple and integrated as a part of iOS 5 in 2011.

It works like this: You ask your phone a question in natural language – speech recognition is done on the phone – the request is sent to Apple’s cloud servers – processed – and then the reply is sent back to your phone.

siriThere is two issues with this process: Privacy and speed. Apple is well known for it’s focus on privacy, so the machine learning is anonymised, and with every new version of iOS more and more of the process to protect privacy and increase the speed has been moved to your phone.

Siri today is incredibly fast – but not available unless you’re connected to the internet. It wouldn’t be a surprise if newer versions of the iPhone will allow siri to run independent of the internet, leaving us to require connection only for communicating with 3.rd parties to book the services.

The trend; led by Apple and Qualcomm, is to build silicon that handle deep learning on the device. Qualcomm is producing the Snapdragon processors powering most of the high end Android phones. In May Qualcomm launched a software developer kit called the “Natural Processing Engine” for its Snapdragon 820 chip, that powers 2016 high end Android phones like Samsung Galaxy S7, Moto Z, OnePlus 3, HTC 10 and many more.

 

Natural AI experience
The Aipoly app launched in January, was not the first seeing app for blind or visually impaired people, but it was the first app that dropped the cloud, saving the critical few seconds it used to take to get the answer. Aipoly recognises objects and speaks their names aloud. It not only recognises that you are pointing at an animal, but which race and even the name if it is a pet you meet over again. With a direct answer you get a more natural AI experience.

Screen Shot 2016-09-26 at 16.22.45You don’t have to be a rocket scientist to imagine how Machine learning on your phone is going to change augmented reality far beyond Pokemon Go.

When your phone is able to map your surroundings in detailed 3D and add accurate animations into your video feed on your screen, augmented reality will become much more real-time and natural.

Also Virtual Reality will start looking more genuine with better mobile AI integration. For VR it is imperative that video and audio is mapped to eye-, head- and gesture-tracking in a flawless manner to create a virtual experience that actually feels real. And there is no doubt that investments in mobile AI silicon is going to set a trend in VR the coming years.

If future AI does not need the cloud, then the cloud will not require your personal data. That is certainly eliminating the biggest “creep-factor” of Artificial intelligence.

 

Mobile games is the reason
There are now 500.000 mobile games in Apple’s AppStore and mobile gaming is expected to gain the number one position in the gaming industry measured by revenue in 2016. Mobile games require top-notch silicon, and have been an important driver for the development of mobile processors like the Apple A10 and Snapdragon 820.

Todays mobile processors performance are very close to mainstream laptop processors and are offering lightening speeds. A few years earlier this was not the case. But as the global smartphone market has become way larger than the PC market, the investments in research and development have done so likewise. The GPU in the iPhone 7 is 240 times faster than the first iPhone launched 9 years ago, the last years improvements of GPU speed is 50% so speed increase is actually accelerating.

Please contact us at Cloudnames if you have any questions. You can send us a direct email on hello@cloudnames.com or simply leave your contact details below and we’ll get right back to you.

Book Free Consultation

Are you falling behind with digital marketing? Don’t lose important customers. If you need help formulating a strategy that works best for you without throwing away your hard earned cash, we can evaluate your current online position and give you expert advice that suits your business goals. Contact us today to speak to a member of our team!

 

Prisma photo app
Another exciting app that has moved from the cloud to device is the Prisma photo app that renders your image in many different artistic styles. An Instagram filter simply tweeks basic parameters like color, brightness and white balance, the Prisma app analyses the image, recognise objects, shapes, lines, colors and shade, and then redraw the image from scratch using Edvard Munch’s or Picasso’s style.

Screen Shot 2016-09-26 at 16.46.04The Prisma app started as a cloud service, but launched the new iPhone app in August purely on device and plan to deliver an Android version on device as well. On the new iPhone app the effects are rendered almost instantaneously while it will take a few seconds with the cloud service. Prisma is also planning an app that does the same for video. Delivering that as a cloud service would have been expensive in terms of server capacity and bandwidth, so all the AI rendering will be done on the phone.

 

Specialised AI chips
Intel just recently announced the purchase of VPU producer Movidius. The latest VPU (Vision CPU) chip “Myriad 2” is among other used in the DJI’s Phantom 4 drone, helping it avoid objects, hover in place and track moving objects like for instance a skier. And it is widely expected that similar AI technology will find its way to mobile devices soon.

In the new iPhone 7, the camera’s ISP (Image Signal Processor) performs more than a 100 billion operations within 25 milliseconds for each photo. “This truly is a supercomputer for photos,” said Phil Schiller at Apple’s launch event. He was referring to the the huge improvements to the new camera system. “For the first time, it has full cinematic color and can take multiple photos and fuse them into one perfect image,” he said.

The ISP reads the scene and uses machine learning to identify objects, people and scenery within it. From there, it sets exposure, focus, wide colour, tone mapping, noise reduction and combines multiple images every time you take a photo.

Apple is notoriously known for keeping secrets about future products and services, so it is no surprise that they have not revealed their plans for future AI initiatives. But in many comments made by the company it seems evident that Apple is moving AI capabilities to the iOS devices and build specialised silicon to support the move and keep personal information private.

 

Facebook and Google
Two other companies that is investing a lot in AI is Facebook and Google. Facebook CEO Mark Zuckerberg presented their 10-year plan in April, revealing that AI is one of three major focus areas, along with connectivity and VR/AR. The first AI service from Facebook was the News Feed which uses machine learning to decide what to show to more than 1,7 billion users. On the F8 developer conference AI chatbots for Messenger was launched.

Google have invested a lot into the AI in the google search, their Adwords platform and the Google Display Network and many other products like Maps, Gmail, Translate etc. This is technology billions of users are exposed to every day based on Artificial Intelligence. But the parent company Alphabet also make huge investments in “future bets” which include biotech, AI, robots, self driving cars, VR and many other technologies requiring AI.

Facebook and Google business plans are largely based on offering free technology in the cloud, collecting information about the users and then placing ads based on that information. Many critical voices are sceptical about the privacy of users when these two giants collect all information they can from their users – and even users that are not using their services.

There are thousands of other companies and startups that invest in Artificial Intelligence, Machine learnings and Neural networks. Who will win the market is for the future to show, but I think we can safely say that the clash between the titans will be fought on your smartphone – for now.

 

What we do at Cloudnames
Cloudnames is a Digital Agency that can help you with everything you need to be your best online. We can manage your complete Digital Strategy, your website, Social media profiles, your Digital Marketing, Sales and Marketing automation and all the complicated technical details to effectively grow your revenue online.

We offer a managed web design service where we manage your website and keep it updated and secure on a regular basis. We can also take care of your Search Engine Optimisation, Search Engine Marketing, Social Media Marketing and all other Digital marketing campaigns you need. (Slightly shortened)

Please contact us at Cloudnames if you have any questions. You can send us a direct email on hello@cloudnames.com or simply leave your contact details below and we’ll get right back to you.

 

Contact Us

At Cloudnames we offer all the services you need to be your best online. We can handle your overall Digital Marketing project, while you focus on your core business. Contact us today, and let us know what you need and we’ll get back to you with real expert advice!

Sharing is Caring!
At Cloudnames our motto is “sharing is caring,” so we have made a guide on everything you need to know to be successful with Digital Marketing. You can download your free copy today (Link to Digital Strategy LP)

Other articles you might like:
– Apple iPhone 7 launch
– Do Robots have feelings?

Please follow our blog if you are interested in Social Media, Digital marketing and technology. What is your favorite AI tool? Please let us know in the comment field below and share with your friends by clicking the icons. Remember sharing is caring!