25 Feb Tech 2019 Predictions -Part 2
Prediction 4: On-Device AI Will Start to Shift the Conversation About Data Privacy -Tech 2019 Predictions
One of the least understood aspects of using tech-based devices, mobile applications, and other cloud-based services is how much of our private, personal data is being shared in the process often without our even knowing it. Over the past year, however, we’ve all started to become painfully aware of how big (and far-reaching) the problem of data privacy is. As a result, there’s been an enormous spotlight placed on data handling practices employed by tech companies.
At the same time, expectations about technology’s ability to personalize these apps and services to meet our specific interests, location, and context have also continued to grow. People want and expect technology to be smarter about them, because it makes the process of using these devices and services faster, more efficient, and more compelling.
The dilemma, of course, is that to enable this customization requires the use of and access to some level of personal data, usage patterns, etc. Up until now, that has typically meant that most any action you take or information you share has been uploaded to some type of cloud-based service, compiled and compared to data from other people, and then used to generate some kind of response that’s sent back down to you. In theory, this gives you the kind of customized and personalized experience you want, but at the cost of your data being shared with a whole host of different companies.
Starting in 2019, more of the data analysis work could start being done directly on devices, without the need to share all of it externally, thanks to the AI-based software and hardware capabilities becoming available on our personal devices. Specifically, the idea of doing on-device AI referencing (and even some basic on-device training) is now becoming a practical reality thanks to work by semiconductor-related companies like Qualcomm, Arm, Intel, Apple, and many others.
What this means is that if app and cloud service providers enable it (and that’s a big if)you could start getting the same level of customization and personalization you’ve become accustomed to, but without having to share your data with the cloud. Of course, it isn’t likely that everyone on the web is going to start doing this all at once (if they do it at all), so inevitably some of your data will still be shared. However, if some of the biggest software and cloud service providers (think Facebook, Google, Twitter, Yelp, etc.) started to enable this, it could start to meaningfully address the legitimate data privacy concerns that have been raised over the last year or so.
Apple, to its credit, started talking about this concept several years back (remember differential privacy?) and already stores things like facial recognition scans and other personally identifiable information only on individuals devices. Over the next year, I expect to see many more hardware and component makers take this to the next level by talking not just about their on-device data security features, but also about how onboard AI can enhance privacy. Lets hope that more software and cloud-service providers enable it as well.
Prediction 5: Tech Industry Regulations in the US Becomes Real – Tech 2019 Predictions
Regardless of whether major social media firms and tech companies enable these onboard AI capabilities or not, its clear to me that we’ve reached a point in the US social consciousness that tech companies managing all this personal data need to be regulated. While Ill be the first to admit that the slow-moving government regulatory process is ill-matched to the rapidly evolving tech industry, that’s still not an excuse for not doing anything. As a result, in 2019, I believe the first government regulations of the tech industry will be put into place, specifically around data privacy and disclosure rules.
Its clear from the backlash that companies like Facebook have been receiving that many consumers are very concerned with how much data has been collected not only about their online activities, but their location, and many other very specific (and very private) aspects of their lives. Despite the companies claims that we gave over most all of this information willingly (thanks to the confusingly worded and never read license agreements), common sense tells us that the vast majority of us did not understand or know how the data was being analyzed and used. Legislators from both parties recognize these concerns, and despite the highly polarized political climate, are likely going to easily agree to some kind of limitations on the type of data thats collected, how its analyzed, and how its ultimately used.
Whether the US builds on Europes GDPR regulations, the privacy laws instated in California last year, or something entirely different remains to be seen, but now that the value and potential impact of personal data has been made clear, theres no doubt we will see laws that control the valued commodity that it is.
Prediction 6: Personal Robotics Will Become an Important New Category – Tech 2019 Predictions
The idea of a sociable robotone that people can have relatively natural interactions withhas been the lore of science fiction for decades. From Lost in Space to Star Wars to WallE and beyond, interactive robotic machines have been the stuff of our creative imagination for some time. In 2019, however, I believe we will start to see more practical implementations of personal robotics devices from a number of major tech vendors.
Amazon, for example, is widely rumored to be working on some type of personal assistant-based robot leveraging their Alexa voice-based digital assistant technology. Exactly what form and what sort of capabilities the device might take are unclear, but some type of mobile (as in, able to move, not small and lightweight!) visual smart display that also offers mechanical capabilities (lifting, carrying, sweeping, etc.) might make sense.
While a number of companies have tried and failed to bring personal robotics to the mainstream in the recent past, I believe a number of technologies and concepts are coming together to make the potential more viable this year. First, from a purely mechanical perspective, the scarily realistic capabilities now exhibited by companies like Boston Dynamics show how far the movement, motion, and environmental awareness capabilities have advanced in the robotics world. In addition, the increasingly conversational and empathetic AI capabilities now being brought to voice-based digital assistants, such as Alexa and Google Assistant, demonstrate how our exchanges with machines are becoming more natural. Finally, the appeal of products like Sonys updated Aibo robotic dog also highlight the willingness that people are starting to show towards interacting with machines in new ways.
In addition, robotics-focused hardware and software development platforms, like Nvidias latest Jetson AGX Xavier board and Isaac software development kit, key advances in computer vision, as well as the growing ecosystem around the open source ROS (Robot Operating System) all underscore the growing body of work being done to enable both commercial and consumer applications of robots in 2019.
Prediction 7: Cloud-Based Services Will Make Operating Systems Irrelevant – Tech 2019 Predictions
People have been incorrectly predicting the death of operating systems and unique platforms for years (including me back in December of 2015), but this time its really (probably!) going to happen. All kidding aside, its becoming increasingly clear as we enter 2019 that cloud-based services are rendering the value of proprietary platforms much less relevant for our day-to-day use. Sure, the initial interface of a device and the means for getting access to applications and data are dependent on the unique vagaries of each tech vendors platform, but the real work (or real play) of what we do on our devices is becoming increasingly separated from the artificial world of operating system user interfaces.
In both the commercial and consumer realms, its now much easier to get access to what it is we want to do, regardless of the underlying platform. On the commercial side, the increasing power of desktop and application virtualization tools from the likes of Citrix and VMWare, as well as moves like Microsofts delivering Windows desktops from the cloud all demonstrate how much simpler it is to run critical business applications on virtually any device. Plus, the growth of private (on-premise), hybrid, and public cloud environments is driving the creation of platform-independent applications that rely on nothing more than a browser to function. Toss in Microsofts decision to leverage the open-source Chromium browser rendering engine for its next version of its Edge browser, and its clear were rapidly moving to a world in which the cloud finally and truly is the platform.
On the consumer side, the rapid growth of platform-independent streaming services is also promoting the disappearance (or at least sublimation) of proprietary operating systems. From Netflix to Spotify to even the game streaming services mentioned in Prediction 2, successful cloud-based services are building most all of their capabilities and intelligence into the cloud and relying less and less on OS-specific apps. In fact, it will be very interesting to see how open and platform agnostic Apple makes its new video streaming service. If they make it too focused on Apple OS-based device users only, they risk having a very small impact (even with their large and well-heeled installed base), particularly given the strength of the competition.
Crossover work and consumer products like Office 365 are also shedding any meaningful ties to specific operating systems and instead are focused on delivering a consistent experience across different operating systems, screen sizes, and device types.
The concept of abstraction goes well beyond the OS level. New software being developed to leverage the wide range of different AI-specific accelerators from vendors like Qualcomm, Intel, and Arm (AI cores in their case) is being written at a high-enough level to allow them to work across a very heterogeneous computing environment. While this might have a modest impact on full performance potential, the flexibility and broad support that this approach enables is well worth it. In fact, its generally true that the more heterogeneous the computing environment grows, the less important operating systems and proprietary platforms become. In 2019, its going to be a very heterogenous computing world, hence my belief that the time for this prediction has finally come.