2017 is promising to become the year of artificial intelligence and machine learning. It started with the CES show in Las Vegas earlier in the year where one could clearly see this new shift in technology. AI is showing up in one fashion or other in the various gadgets that we use, be it electronics or electric cars.
- 1 Amazon and Microsoft Up Their Game
- 2 What About Apple’s Artificial Intelligence Push?
- 3 Artificial Intelligence– Much More than Virtual Assistants in Future
- 4 Apple’s In-House Talent Leadership in AI and Machine Learning
- 5 Writing off Apple? Think Different
Amazon and Microsoft Up Their Game
The CES show highlights Amazon’s expanding influence when it comes to AI. Alexa from Amazon is quickly establishing itself as a leading brand when it comes to adoption of AI centric products.As Amazon rolls out new Alexa partnerships, Microsoft is also stepping up its game in this area. The company announced that it acquired a Canadian based startup called Maluuba this month. With this new AI based acquisition, Microsoft is planning to take AI to the masses. In its’ announcement this week, Maluuba noted:
“Microsoft is an excellent match for our company. Their ambitious vision of democratizing AI to empower every person and every organization on the planet fundamentally aligns with how we see our technology being used.”
This new acquisition expects to bring more power to Microsoft products such as Cortana as well as offer AI-based solutions as a part of Microsoft’s enterprise offerings. Google’s Home product launch showcased the company’s interest in commercializing machine learning via its new virtual assistant, and the company has been spending millions in marketing both the Pixel phone and the Google Home.
What About Apple’s Artificial Intelligence Push?
Although Apple doesn’t attend the CES shows or publish press guidelines around upcoming products and features before they are released, Tim Cook has made it clear that AI and deep learning technologies are a big focus moving forward.
Naysayers are plenty when it comes to Apple, and they do not waste time in suggesting that Apple is so yesterday when it comes to innovation in this exciting area.
I beg to differ and would like to put forth my rationale:
Apple doesn’t have a Home Speaker Device. Does it Really Matter?
In the case of Amazon, the Echo home speaker allows the company to position its AI technology straight inside your home. The same goes for Google. The Google Home product allows the company to get a share of your household space and position itself as a virtual assistant.
Why doesn’t Apple have a similar device? Many Apple fans have speculated that Apple could, in fact, introduce a product similar to Echo and Google Home this year and position itself in this new evolving product category.
I am not sure if a new home speaker from Apple with Siri capabilities really adds a ton of value, either for the customer or the company. After all, how many devices do you intend on carrying from one room to the next in your household to access your virtual assistant? You can practically do almost all the things with Siri today on your iPhone or iPad that you can do using the Amazon or Google device. By using Siri on your Apple TV, you can even automate your home via the Homekit app.
Apple is also bringing the Sirikit to Apple Watch with WatchOS 3.2, that will allow third-party developers to build Siri-based apps for the Watch. At the end of the day, It’s the power of Siri that will determine if Apple is ahead of the game or playing catch up, not an external speaker that you can carry from one room to the other to use the virtual assistant.
Accuracy and Latency, Key Litmus Criteria
In my opinion as an Apple fan, Apple is following the right strategy when it comes to pursuing AI and machine learning within Siri and beyond Siri.
There are two key elements for success when it comes to AI based product offerings. The first is the accuracy of the results from the algorithms that provide you the answer to your queries and the second is the latency associated with the responses.
Apple fares very well when you compare Siri’s performance with that of other voice assistants. Siri not only beats out both Google and Amazon when it comes to accuracy of results but is also ranked higher than many disrupting startups in this field.
When it comes to the latency of the results, you can clearly test it yourself and see that the responses from Siri are lighting fast (Although I must admit it can be frustrating to find the right results from time to time).
The accuracy of results is also based on how much data the company has for its algorithms to learn and provide meaningful results. In this area, Apple has a lead over its competitors since its Siri servers have been collecting information for years and figuring out ways to make it more accurate.
Scaling via Complementary Technologies and Acquisitions
It is no accident that Apple has opened up Siri to third-party developers with iOS 10. It is clearly a part of a very elaborate strategic roadmap that the company is pursuing to establish Siri as a leader.
Apple’s services revenue has been growing by leaps and bounds over the last few years. Its execution of partnership strategies with independent third party developers has yielded success when it comes to generating top line revenue via the AppStore.
App developers earned $20 billion in 2016, up 40% from 2015.
This is where some of its AI acquisitions may complement its success going forward. Provide some of the machine learning and AI power to the third party developers and it could spur some innovation around new use cases.
Apple’s acquisitions of Turi and Tuplejump in 2016 suggest that the company is exploring new and innovative approaches to beef up its AI capabilities.
In 2015, the company acquired smaller startups in this emerging area that included Perceptio and VocalIQ. The company has also invested heavily in the new Softbank offering to seek out new technologies.
How could these acquisitions potentially help Apple?
Let’s look at a few of these and see what potential value they can bring to the table.
Turi lets developers build apps with machine learning and artificial intelligence capabilities that automatically scale and tune. Its products are principally designed to help large and small organizations make better sense of data.
Use cases include recommendation engines, fraud detection, predicting customer churn, sentiment analysis, and customer segmentation.
One could infer that Apple may be able to provide some of Turi’s capabilities and tools in the future to third party app developers to make it easier for them to expand on AI capabilities to Siri and Apple’s iMessage platform.
Apple has been investing heavily into both Siri and iMessage over the past few years. It introduced significant changes to iMessage with iOS 10 and introduced Siri on its macOS platform in 2016. Now, it is positioning Sirikit on the Apple Watch with watchOS 3.2. If the company can make it easier for third-party developers to seed some of the emerging AI capabilities via Sirikit, it can easily position itself as a leader. The large third party developer base can become a key catalyst when it comes to commercializing some of the AI features via Siri.
Privacy and AI for Mobile platforms
For AI based platforms to work, they need huge volumes of data that they can parse and tag quickly. This would require that users are willing to share more data with the AI platforms to get accurate results. The problem is privacy.
How much data are you willing to share to get more accurate and actionable results from the AI Platforms? This is where it makes sense to have a platform that can offer accurate and meaningful results while mitigating some of the data privacy issues. Apple has been a leader in this area and has championed user privacy when it showcased Differential Privacy last year.
Its acquisition of Perceptio is a signal that it intends to pursue a well-balanced strategy between data privacy and AI capabilities. Perceptio is one of the few startups that build a platform which allowed companies to run advanced artificial intelligence systems on smartphones without needing to share as much user data. This 2015 AI acquisition fits Apple’s strategy of trying to minimize its usage of customer data and do as much more processing as possible on the device.
Artificial Intelligence– Much More than Virtual Assistants in Future
Today, when we think about AI and machine learning, the first application that comes to our mind is Voice Assistants.
AI is much more than that. Intelligent Image processing is an area, which is expected to grow as users move away from text-based systems to image/emoji based communication. We have seen some nascent AI offerings in iOS 10 Photos. Apple is building up its capabilities in this area as well.
One of the small acquisitions that the company made was Emotient. This is a corporation that uses AI to recognize and act upon facial expressions. As camera system offering such as Dual lens become standard features of smartphones, we are bound to see more intelligent image processing based features from Apple. It wouldn’t be surprising to see a picture based security feature on the next iPhone.
Apple’s leadership in positioning voice assistant systems in cars is not new. Apple introduced Siri Eyes Free concept way back in 2013. The Ferrari FF model rolled out a full version of CarPlay in Sept 2014.
It surprises me when some people make a big deal out of the latest Alexa/Ford partnership. Carplay has been around for some time now, and today you can even use your Apple Watch to control your TESLA functions.
Healthcare is another area where Apple is slowly establishing its leadership by offering CareKit and HealthKit and working with large pharma companies and other research institutions. This is one area, which is bound to benefit from the use of AI and machine learning technologies. As more sensors get integrated into wearables or other medical devices, there will be a greater need for a platform that can synthesize massive volumes of sensor data and produce meaningful insights.
This is where Tuplejump can complement Apple’s machine learning technologies. This 2016 acquisition provides Apple with a platform to process massive amounts of various sensor data in an easy and timely fashion.
Apple’s In-House Talent Leadership in AI and Machine Learning
Over the years, Apple has been very successful in building up a stellar team filled with some of the who’s who when it comes to AI, Machine Learning, and Siri.
Rusian Salakhutdinov is no stranger when it comes to AI. He was a high-profile hire from Carnegie Mellon last year and currently is the Director of AI research at Apple. With prior experience at Samsung, he understands some of the challenges when it comes to AI and smartphones.
Tapani Raiko, the former co-founder of Curious AI Company, has been working in AI research at Apple since mid-2016. He’s an industry expert when it comes to unsupervised machine learning. As the nascent machine learning technology matures into more unsupervised applications, Apple is bound to have an edge with some of its in-house talent who work alongside Tapani.
Nicolas Pinto, former head of Perceptio is also actively engaged in AI initiatives at Apple. A deep learning ninja and AI evangelist, he has worked with DARPA and Google Research efforts in the area of computer vision and computational theories around how visual cortex accomplishes object recognition.
It’s no surprise that at the WWDC 2016, Apple introduced not one, but two neural network APIs, called Basic Neural Network Subroutines (BNNS) and Convolutional Neural Networks (CNN) and made it available to developers. With team leaders who have not only a terrific academic research credential but also industry experience, Apple has positioned itself very well when it comes to securing talent to build out the future with some of these exciting technologies.
Writing off Apple? Think Different
There is no denying the fact that iPhone sales have slowed over the past few years. This is largely a function of the prevailing economic conditions as well as a lack of innovative iPhones. Hardware improvements via OLED Screens or new camera systems can only go so far in convincing users to upgrade their smartphones. It’s the software features, value for privacy and new capabilities that will position Apple’s leadership in the future.
During the September 2016 Earnings call, Tim Cook remarked:
“As you know, iPhone customers are the most satisfied and loyal customers in the world, and fiscal 2016 saw more customers switch from Android to iPhone than ever before. This is due to the superior customer experience we deliver with our products, and it’s something no other company can match.”
It will be interesting to see how this trend evolves over the next few quarters from Apple’s earnings.
If Apple can deliver AI and machine learning capabilities by using the intellectual capital from its acquisitions and its in-home talent within a reasonable timeframe, it is bound to create renewed excitement among its fans. Apple has recently announced plans of joining the AI Partnership along with Google and Facebook
By no means, you can write off Siri just because Microsoft made a new machine learning acquisition or Amazon forged a partnership with Ford to offer Alexa in vehicles. Good Things Come to those who wait, and as an Apple fan, I’m willing to wait and see what the 10th anniversary of iPhone has in store for us in 2017.
Obsessed with tech since the early arrival of A/UX on Apple, Sudz (SK) is responsible for the editorial direction of AppleToolBox. He is based out of Los Angeles, CA.
Sudz specializes in covering all things macOS, having reviewed dozens of OS X and macOS developments over the years.
In a former life, Sudz worked helping Fortune 100 companies with their technology and business transformation aspirations.