AI Integration Redefines Mobile Cloud – Personalization


The CEO of Nvidia, Jen-Hsun Huang, recently blogged about the new “AI computing era,” which, as he put it, is succeeding the prior great wave in computing – “the Mobile-Cloud era.” With AI taking hold, and expanding like “wildfire,” we’re quite interested in its growth and relationship to the Mobile Cloud.

Ultimately, we believe, this relationship will focus largely on “personalization” – a Mobile Cloud objective that is fostered by AI. However, in many other respects, AI can also be viewed as leading to an extension of the Mobile Cloud. AI for the masses will have an essential symbiotic relationship to the cloud.

Catching Up on AI’s Growth

There has been phenomenal activity and progress in AI. Huang points out that, just regarding the usage of Nvidia’s key technology GPU (graphics processing units), “the number of GPU deep learning developers has leapt 25 times in just two years.” There have been some 1,500 AI startups.

AI, Artificial Intelligence, refers to the capability of machines to exhibit intelligent functions of learning and problem solving, “cognitive” functions resembling the human mind. To keep things simple, we will speak only of AI, whereas others may describe the phenomenon, or parts of it, by other names – such as “cognitive computing” (e.g., IBM’s CEO refers to the new era as “the cognitive era”), “machine learning,” “deep learning,” “neural networks” – each of which may have some different nuances.

Developments in AI are pouring out of myriad sources. One sign of the times is that, in the area of Autonomous Vehicles, where there has been a proliferation of developments almost weekly, in October, Uber-owned Otto, and Budweiser, cooperated on an autonomous real-life, beer truck delivery in Colorado, a 120-mile trip (a human driver assisted with on- and off-ramp navigation.)

In healthcare, researchers, for example, are pushing ahead with determining how robotic arms could assist doctors and nurses, even in operating rooms. In retailing, in November IBM has acquired the Expert Personal Shopper (XPS) technology from Fluid, a Watson-based AI platform for providing shoppers with personalized information in a dialogue mode.

A good deal has been written about the status of AI on the enterprise front. Among the most interesting insights, one observer states that acceptance of AI or machine intelligence at the corporate level is almost a binary phenomenon – some people “get it” and most others don’t, to date. The idea of “machine trust is daunting.” (“The Current State of Machine Intelligence 3.0,”

AI & Mobile Devices – Viv & Personalization

The area of incorporating AI into mobile devices is clearly heating up. Pioneered by Apple with Siri, the most notable recent addition was the Google Pixel device, which included Google’s Assistant technology. Google touts this as creating a “personal Google” for a user, pointing out that Assistant can not only act as a voice-controlled search adjunct, but also can remember specific things about the individual, such as personal preferences.

In addition, in October Samsung announced the acquisition of AI innovator, Viv, which was founded by the developers of Siri. Viv reportedly has a number of developments that could extend the range of the mobile device personal assistant. Viv management lists the key functions and criteria as: 1) providing a single assistant; 2) personalized for the individual; 3) available on any device; 4) powered by every service. Regarding this last point, Viv asserts that it has been building a “third party ecosystem” which will be a crucial factor in success of the personal assistant. The company characterizes its rising body of work as “conversational commerce.”

In a presentation earlier in May 2016, Dag Kittlaus, co-founder of Viv explained that Viv uses speech recognition from Nuance. Once it recognizes the “intent” of the user’s request, it works its “magic,” which the company calls “dynamic program generation.” In other words, the system writes a program by itself, which can apply to fulfilling requests across any information domain or topic.

Kittlaus expressed a vision of cloud-based AI agents that could work on any type of device and would virtually “follow the individual around.” He explained personalization as, essentially, relieving the user of the need to explain things, and asserted, “The rise of assistants is inevitable.”

Our Take

We expect the advancement of AI – anticipated in fact and fiction since the dawn of the computer business – to be inexorable.

Regarding the future of Mobile Cloud, a great number of areas of AI application will basically draw on the convergence of these two factors. Robots, whether in factories, hospitals, hotels or elsewhere are examples of extending the Mobile Cloud. Of course, there will be some AI applications that are clearly fixed cloud in nature, such as, for example, monitoring and maintenance of giant turbines.

However, one of the transcendently important areas of application of AI is clearly that of mobile devices. Our take is that mobile devices, which have been advancing for about the last ten years, are barely scratching the surface of their future potential for society.

We expect to see major reformulations and disruption to the very nature of mobile devices even in the next two-to-three years. We expect AI to be an integral part of this redefining of the mobile device.

In the area of personal assistants, we believe the mobile industry is at an extremely early stage. While developments like Viv are innovative, our impression from their presentation was that while they appear to be accumulating contextual experience as to user needs, the system is still reactive in nature, and their solutions still may appear to go, at this point, to pre-selected sources.

In any case, we believe the potential for personal assistants is for a much more robust and active engagement with the individual users. So much of what’s being shown today appears to be still looking like glorified Search, with a voice front end.

The Big Monsters of mobile – Apple, Samsung, Google, Huawei – all have their eyes on exploiting AI in future generations of devices. It’s been noted that promising early stage AI companies have been voraciously snapped up by major IT players. For example, Google and DeepMind, Apple and Siri as well as Turi, Samsung and Viv. (Huawei which has announced a strategic emphasis on AI recently invested $1 million in an AI partnership with US Berkeley.)

It’s interesting that with all of the resources of the big guys, at almost every turn they’ve been forced to acquire expertise in AI from near-startups and other emerging companies. We’ve been anticipating the personal assistant for over 20 years now. Frankly, we believe that more agile companies, many yet to emerge, that are more devoted to AI, will probably move us closer to realization of the vision in the future before the Big Monsters do.

There are innumerable other opportunities for the application of AI in businesses and in virtually all other aspects of life. However, one aspect of AI that we find fascinating is its relationship to Big Data.

The term Big Data is of far more recent vintage than the concept of AI, Big Data having been popularized, within about the last five years. Three years ago, we wrote the following regarding the problems raised by Big Data:

“Big Data is another term that has gained popularity in the past two years. The term, in our view, is not very helpful. The amount of data grows over time regardless, and while there will be an explosion of data as the number of endpoints that can communicate information about themselves (e.g., M2M) expands, it isn’t really the amount of data that is important, but the need for facilities and tools to store and manage it, and, even more so, the analytic capabilities to process that data, so that it doesn’t become a rather useless attic full of information.”
(“The Mobile Cloud: Market and Outlook to 2017,” BSG Advisory LLC, 2013)

Fortunately, AI has been progressing in leaps and bounds and one of the leaps has been to the rescue of Big Data. Without AI Big Data would threaten to drown the world in a morass of trivia and irrelevancy.

Visit related links: