It's hard to have a conversation about enterprise technology without the subject of artificial intelligence (AI) being a focal point. Recent Ovum research shows that 71% of enterprises are either considering or planning to use AI, bots, and machine learning over the next 12 months (Ovum's 2017/18 ICT Enterprise Insights survey), and 65% of organizations believe AI will have an impact on their workplace over the next three to five years (Ovum Workspace Services Survey, 2017). As mobility is such a vital digital workplace element, it is important to understand how AI capabilities will help further empower the mobile workforce going forward.
AI can be very processor-intensive, and to overcome this issue, some device manufacturers are introducing dedicated AI processors into their mobile devices, with Huawei being one such example. The company recently introduced a dedicated AI chip called a Kirin 970 into some of its recently released smartphones. Having a dedicated chip to handle AI processes means that device power can be reserved for other tasks that a user may wish to undertake on the device, improving the user experience as a result. This dedicated hardware capability means AI processing can take place on the device itself, with no reliance on the cloud or app, as is often the case with many AI applications.
Dedicated AI chips open up new possibilities in how mobile devices can be used within organizations. A good example is in how these chips can help speed up image recognition and processing. When powered by AI and machine learning capabilities, smartphone cameras can add another layer of business security and authentication through functionality such as face ID and behavioral analysis. Cameras can also be used to profile an environment via visual scanning, providing people with a greater level of real-time information not only on the environments they are working in, but on things and objects within that environment. Samsung Bixby, for example, enables people to learn about items that can be scanned via a smartphone camera. These scanned items could potentially be anything, and they need not be connected to the internet. This camera use case also shows how AI and machine learning, in combination with other capabilities, can help businesses improve user experiences and the process and practices that support them.
An important benefit often touted by hardware manufacturers relating to on-device AI chips is that they are not dependent on any network capability – all tasks and processing can be handled on the device itself. However, it is important to note that many enterprise application providers, especially those that have platforms built in the cloud, are developing their AI capabilities to be network dependent, as this is the way such applications are commonly used anyway. Because many enterprise apps are now hosted and utilized via the cloud, this lack of dependency on the network may not in fact be a feature that many employees will see great benefit from day-to-day, as many employees work online and with apps that offer AI capabilities that have been developed to always be online. Developers do benefit, however, as they have flexibility in whether they create apps leveraging on-device AI capabilities, those in the cloud, or a combination of both.
Adam Holtby, Senior Analyst, Workspace Services