If you’ve spent any time playing with famed Apple digital assistant, Siri (especially its spiffed-up iOS 7 version), you’ve probably had some sort of wow experience with the software. A moment, however short, that made you think, “Okay, this thing is actually pretty impressive.” As a presumed tech enthusiast, you’ve probably also heard of Watson, the supercomputer that dominated several rounds of Jeopardy! against some seriously smart human competition.
Now imagine having Watson-like intelligence on your smartphone or tablet, bringing the answers to your questions and searches at speeds that make Siri seem like the Dewey Decimal System. That’s the aim of a new remote-access service called Watson Developers Cloud. When the first apps using the service hit smartphones and other devices in 2014, you can expect an explosion in several tech-related trends — cloud computing, macro and micro data collection and personalized software-based services — to follow in short order.
Different though they may be, Siri and Watson Developers Cloud share a lot of similar ideas on the back end. Both utilize the user’s hardware as a sort of gateway to larger, more powerful computers in a remote location, receiving and resolving the user’s requests over the Internet much more effectively than their device may have been able to do locally. They both flaunt impressive pedigrees as well, with Watson coming from tech giant IBM and Siri tracing its roots back to a privatized DARPA project.
The big difference between the two lies in their respective intended user bases. Outside of smallish, specialized initiatives like Siri Eyes Free, Apple’s focus has largely been on their mobile customers. Watson Developers Cloud, on the other hand, is built from the name up for software developers who “need only have their applications make a REST API call” to harness whatever capabilities IBM sees fit to share, according to a recent Gigaom.com update.
Is it a totally level comparison? Probably not, considering who each product is designed for. Their above-mentioned similarities and computing platform of choice (namely mobile) could still result in direct competition. Some news outlets have been predicting a fight for the ages. One Forbes article called Apple’s product “David” and IBM’s “Goliath” in reference to what it saw as an upcoming battle for supremacy in the world of mobile health (mHealth). Perhaps even more telling are the strengths and weaknesses the Forbes article lists for each platform:
“Siri may not have as much computing power, but the application is inexpensive, easy to use and travels just about anywhere.”
Even in its early stages — the initiative was only announced in November — it’s clear that Watson Developers Cloud has a chance to address all those potential problems, making it a likely game-changer for any app with features remotely resembling information retrieval or smart customization.
Telling Siri to look up a question on Wolfram Alpha is cool, but asking it very specific and complex questions about Isaac Newton, then receiving a correct answer from the software itself — something that happened during the above-mentioned Jeopardy! thumping — would be absolutely jaw-dropping.
Although there are several other digital assistants readily available for your smartphone, Watson still has plenty of potential value. This is mostly due to one of the supercomputer’s other strengths, learning, as well as the nature of cloud-based computing as a whole.
Computers have been guessing at human preferences for a long time, with varying levels of success. One widely known tale about a Minnesota Target store finding out a teenager was pregnant before her father did illustrates the dumb and smart sides of automated personalization: The system was good enough to root out a person belonging to a profitable demographic, but not good enough to stop itself from spoiling the surprise.
Watson doesn’t have any more tact than your average cash register, but its ability to learn from its mistakes (not to mention the team of IBM engineers backing it up) would all but ensure that it didn’t make the same goof twice. Imagine that kind of self-correcting behavior applied to, say, your Netflix queue.
If you left your account logged in at a friend’s house and he spent the entire next day watching Saturday morning cartoons, a Watson-powered system would potentially be smart enough to tell itself the binge was a fluke. From there, it could potentially check against your real viewing habits to make sure you didn’t want an occasional episode of Rocko’s Modern Life alongside your historical dramas.
That level of customization is good for more than entertainment or convenience. As with Target’s attempts to learn about individual customers, a Watson-powered app could be used to nail down a shopper’s preferences and build customized (and highly accurate/relevant) shopping lists based on any number of user-defined criteria.
Users could enter their weekly desired menu and have the Watson-powered app search for all the individual ingredients, tell it to look for the cheapest version of stuff they usually buy or ask it how to cut calories from their favorite meals and base its findings on that. The same thing applies to travel, dining out, home/auto repair, job searching… the list goes on.
Then there’s healthcare. Watson’s strengths as a supercomputer in the field of mHealth have already been well documented. As a smartphone app, especially one tailored to the medical community, the effects could be world-changing. Two Watson-powered apps due in 2014 — one related to personal well-being and another designed to help healthcare professionals make diagnoses — will undoubtedly be the first of many health-specialized mobile software offerings unlike anything we’ve ever seen.
Watson isn’t the only player in the cloud-powered supercomputer game. Amazon, another tech giant known for its own big-time tech innovations (delivery drones, anyone?), offers a somewhat similar concept with its Elastic Compute Cloud (EC2) service. Yet another major technology player, Google, sells supercomputer processing time via its Cloud Platform.
On the small scale, services like OnLive allow gamers and others to stream high-end software from computers other than their own, while the previously covered Ubuntu for Phones uses similar software to stream virtualized Windows apps to the Linux-based mobile OS.
The difference here is functionality. For all the specialized supercomputer hardware out there — and all the countless virtual computing solutions available to developers and end users — Watson is the only supercomputer with developer-focused cloud connectivity to offer such an extreme level of consumer-friendly technology.
Because the computer is built with cognitive computing in mind, not straight-up processing power, it comes with a number of features other supercomputers don’t have or need, such as voice recognition and the ability to answer questions in natural language. That’s not to say Google won’t announce some extra-smart version of Google Now, but at the moment, Watson is one of the only kids on the block when it comes to cloud-based cognitive computing for end users. While it could compete with a gazillion different apps in a year’s time, for now Watson stands on its own in terms of what it offers, both technologically and on the developer access front.
Revolutionizing the power behind smartphones no longer requires the hardware to be on-site. Watson Developers Cloud (and eventual competitors on the mobile side) can do much of the heavy lifting for their smaller, more portable counterparts. They’ll undoubtedly bring a lot of amazing features along when the first supercomputer-powered apps hit the market and open their doors to developers. No matter what happens in 2014, it’s certainly looking like an exciting time to be a techie.