Imagine if Google Assistant was vastly smarter, collected and stored a wealth of data points about you, and was able to sync your data to any device you'll be interfacing with during the day before you even do so, and you've got a good idea of the potential of Watson Assistant, an assistive AI framework that IBM announced today at its Think 2018 event. That summation may sound hyperbolic, but that is exactly what IBM is selling, and the company is already signing up a range of partners to help it get the technology off the ground and into a range of consumer-facing devices.
One of the most prominent applications for Watson Assistant is I*VIE, shown off in the video embedded below. Made by UK Internet of Things firm Chameleon Technology, the reveal video shows the pervasive assistant essentially following its masters around from device to device, spouting off useful information before it's asked and answering rapid-fire queries and commands with no hesitation or issue. As a bonus, I*VIE even controls power usage and other factors in home devices to help save its master money, and reports how much has been saved each day. Airwire, Munich Airport, Kaon Media, The Royal Bank of Scotland, and HARMAN are all on board, with HARMAN promising to include Watson Assistant in the cockpit of the 2018 Maserati GranCabrio. IFTTT is also signed up, which could very well mean that Watson Assistant's capabilities will be limited only by the imaginations and talents of coders creating applications for it.
IBM's vision for Watson Assistant is vast indeed; the company wants to see it become the new face of IoT, and even stated that competitors like Google and Amazon have actually "laid the groundwork" for Watson Assistant. The bold statement is not entirely without merit; smart devices in homes and smart assistants on phones have made people more willing than ever to hand over information to AI assistants and trust them with daily tasks, two core parts that Watson Assistant will depend on to do its job. Watson Assistant's main attraction is that it learns about not just a user's habits and core information, but about their values, likes, and dislikes, in order to form a comprehensive profile that will help it to determine what's important to each user. Naturally, all of this will follow users from device to device, so long as those devices are integrated with Watson Assistant. It is fascinating that the technology is apparently being revealed so late its development; many next-generation AI that are being revealed and offered to users are still learning how to do their jobs, and users are even told such. Nigel Artificial General Intelligence, still in beta for Android, is a good example of this sort of AI program.