The future of technology is interconnectivity-focused software solutions and A.I. providing a centralized hub for all of the technology that consumers want. That's according to Andy Rubin, Founder and CEO at Essential and the creator of the Android ecosystem, who recently spoke with Time about his own vision on the matter. More to the point, Rubin sees solutions such as the company's up-and-coming Ambient OS as a start in finding a way to bridge the gap between devices and provide consumers with the freedom to choose what technologies they use and when without having to pay so much attention to how well those interact with one another. He also points to what he sees as a need to modernize the way new technologies such as A.I. are being developed to begin with, if technology in general is going to be able to continue advancing at the current rate or better.
Although Rubin's vision for the world of technology, particularly with regard to Ambient OS, could easily be tied to the rise of A.I.-powered digital assistants, Rubin is quick to point out that his idea for a solution should not be seen as a new iteration of those. Alongside the progress of the Internet of Things (IoT), the underlying solution for consumers should be about choice, rather than the creation a new ecosystem for consumers to choose from. The unaddressed problem in the industry is, according to Rubin, the lack of something to bring all of the consumers' electronics together, allowing them to choose when and how they use their devices. He points to the fact that carriers support multiple mobile devices on the same phone number, which frees them up to have more than one device, with each catering to the separate needs of work, home life, and other scenarios. For example, a user can have a more simplified mobile device for convenience when going out on the town or for a jog. That same user can have a larger, more complex device for when they need to work. The next big tech solution, according to Rubin, is in the creation of a software solution to manage those kinds of scenarios and others invisibly without burdening users.
Beyond that, the ultimate goal is to create a customizable experience through the OS that extends across all applications and incorporates situational awareness driven by machine learning. Rubin points out a significant problem with how assistants currently work, both for input and output. For starters, Rubin seems to think that talking to a software-generated person in any public scenario isn't a solution to a problem that exists. Instead, devices should be "completely transparent" to how a user gets information. There are situations where talking to a device can be useful, but those mostly occur when a user is by themselves. On the other side of that equation, Rubin says there really isn't any reason a user should ask for driving directions only to hear directions being given before that user even gets in their car. Furthermore, the issue extends into text-based assistant features. Rubin provides an example by pointing to Gmail's auto-response options, which were introduced earlier this year. He says those aren't nearly diverse enough, and that they should use A.I. to adjust based on the user's own speech patterns and the context of the email – in addition to taking into account things like which device is being used and how that device is generally utilized by the user. Because of those issues, Rubin currently holds the opinion that assistants are already "yesterday's technology."
However, it isn't just Gmail, Rubin says, that changes to how that A.I. functions should be prevalent across all apps. Although he didn't go into great detail, since it's something that the company is working on, it is a solution Essential wants to begin approaching with Ambient OS – which will appear first in the company's Home connected hub. The operating system, according to Rubin, makes progress by being a "horizontal" OS for IoT in much the same ways Android was when it first released. Android, he continues, allowed applications to be run on a wider variety of devices which had previously all had their own operating systems. He also admits that the current market for digital assistants, which is the company's primary focal point for the initial release of Ambient OS, is extremely competitive. Each company working in the space is pouring resources and time into developing its own solutions to the various problems users face. Essential will also have its own Essential Assistant to add to that mix. That could feasibly create problems for the company in terms of getting other companies to allow the use of their APIs under Ambient OS. However, Rubin says that large corporations most likely understand that would be "anti-customer" and is betting that there won't be many issues.
Of course, as Rubin indicates, there are also changes that need to happen in terms of development approaches in order for the industry as a whole to move forward. Because A.I. is data-driven, the systems really need access to a lot of customer information. That's controversial, to begin with, which Rubin says creates a "chicken or egg" problem. Using the example provided by self-driving cars, he continues by explaining that automakers need more data to push for high levels of accuracy and safety, but needs to put more vehicles on the road to collect that data. Both Tesla Motors and Essential, he says, show how that problem can be overcome. In both cases, the respective company focused first on creating a hardware platform and product that performed well in its primary function and incorporated future-proofing features. The focus was then shifted toward a more long-term goal through software, as data was collected. With regard to Essential, Rubin provides the specific example of the device's camera. When the device was first provided to reviewers and users, there were several issues with the camera. Those were able to be fixed with software and, since the company was focused on innovating via software-driven solutions that combine the better light capture capabilities of black-and-white photography with modern color capturing sensors, the camera should continue to see improvements for at least the next couple of years. That same "iterative approach," Rubin says, really needs to be applied at a larger scale across the industry.