UPDATED 16:28 EDT / DECEMBER 05 2011

Siri’s Closest Competition May Come from Detroit, Not Redmond

I make no secret of the fact that I’m a fan of AI and robotics on this blog.  I have only a few dozen feeds in my collection, though, that are totally devoted to the topic of AI, semantics and robotics, unfortunately, because there are very few folks out there that totally throw themselves into coverage of that narrow subset of science and technology.

That changed last weekend, when over the Thanksgiving holiday, it seemed just about everyone suddenly became an expert on semantic analysis, artificial intelligence, and exactly what was going on in the virtual brain of one AI in particular: Apple’s Siri.

Even SiliconANGLE’s own Alex Williams posted a quick hit to his ServicesANGLE column, showing the widely viewed video comparison of Apple’s Siri versus Microsoft’s TellMe.

I spoke with many friends about the coverage, and one prominent pundit known for his Apple fandom told me privately that he found the comparison to be very inaccurate.

“Cartwright’s test was very one-sided (and I’m a Siri fan),” he told me. “For example he didn’t show that TellMe can launch any app on the phone, Siri cannot.”

The way TellMe structures it’s verbal menus is fundamentally different from Apple (and, frankly, serves a quite different purpose). Siri was meant to be a close approximation of a virtual assistant, and as such, has what’s called a “flat grammar” menu system.

TellMe, on the other hand, isn’t always a flat-grammar system (depending on the implementation, which is much different on, say, the TV than it is on the PC or the phone). It also more closely represents a desktop paradigm translated to an auditory realm than a voice-first driven system.

For instance, with TellMe, a query on TellMe sounds like “Bing: Dallas Italian Restaurants,” if you’re looking for a place to eat in town. In Siri, it would sound more like “I’m hungry for Italian.”

Nuances of interface aside, I had to wonder whether or not the general tech punditry hadn’t been the victim of an epic head fake.

Apple’s Siri versus Ford’s Sync

I started thinking about Ford Sync being a more valid comparison point to the Siri product than TellMe shortly after I viewed the video and started polling for the best person to talk to at Ford about the angle, but before I got very far with that, I learned that Ford Motor Company CTO Paul Mascarenas, serendipitous enough, would be dropping by the SiliconANGLE Dallas office to meet with Cali Lewis and the Livid Lobster team.

He was gracious enough to take some time out of his schedule to sit with me and discuss the angle I had come up with: that it was more of a comparable product to Siri than perhaps Microsoft’s TellMe product.

There are a number of surface similarities he and I discussed. Siri and Sync both use Nuance as for the voice recognition end of the product. Both Siri and Sync use natural language processing (otherwise known as semantic processing) techniques to derive the users’ intents. Both, as I mentioned before, use a “flat grammar” system, rather than a hierarchical system. Siri and Sync contain a comparable number of commands; the permutations and combinations can produce a nigh-infinite number of commands, and Sync boasts 10,000 commands after the “Gen 2” release.

Beyond the surface similarities, the two systems diverge a bit. As with anything concerning artificial intelligence, context is everything, and while Siri may be used in a car (or at the office, or in the restroom, or anywhere else), Sync must be used in the car. This allows Sync’s developers to be more focused on the range of commands and utility of the product than Siri’s developers, and perhaps deliver a more polished user experience. In other words, since Sync is geared towards a driver-centric experience, limiting the range of utility to driver-focused queries rather than the limitless queries faced by TellMe and Siri allows them to excel at what they attempt.

Not sure you agree with that sentiment? For proof, you need to look no further than the hub-bub generated by last week’s discovery that Siri had a hard time finding abortion clinics (which, as it turned out, wasn’t because of inherent sexism built into the AI, but because most abortion clinics prefer to go by the more politically correct moniker “family planning clinics”).

BeyondAI: What the Connect Car Means for Mechanical Engineering

In the time we had to speak, Paul and I moved on from the topic of AI into the further reaching implications of what the Sync innovations mean for the future of automotive engineering.

“It’s an shift in thinking for us,” said Mascarenas. “Our customers are used to us shipping a vehicle that’s perfect as soon as it comes out, and will last for 100,000 miles and ten years.”

The shift now, of course, being that Ford is shifting to the mode of development more familiar to us in the tech industry of shipping and iterating. It’s a bit reminiscent of an anecdote (often falsely attributed as direct confrontation between the CEOs of GM and Microsoft) that first surfaced for me in the very early days of the web:

There’s word in business circles that the computer industry likes to measure itself against the Big Three auto-makers. The comparison goes this way: If automotive technology had kept pace with Silicon Valley, motorists could buy a V-32 engine that goes 10,000 m.p.h. or a 30-pound car that gets 1,000 miles to the gallon — either one at a sticker price of less than $ 50.Detroit’s response: “OK. But who would want a car that crashes twice a day?”

Obviously, Ford intends to avoid living up to the “crash early and often” issues that early versions of GUI operating systems saw, but it does open up a world of new possibilities.

Perhaps the most interesting illustration of what the future may bring came when I described to Mascarenas the type of technology coverage we engage in at SiliconANGLE, where I mentioned we attended trade shows like SAPphire and Oracle Open World as well as Hadoop World and Strata.

I knew those weren’t conferences exactly in their wheelhouses, but I encouraged him to drop by our broadcast booth should he be in attendance at one of those shows.

“Those aren’t in our wheelhouse yet, but they very soon will be,” Mascarenas told me. “One of the more interesting possibilities of having an always on connection in our vehicles is the prospect of opt-in data collection. Sure, there are diagnostic and early problem detection possibilities that happen when you look at the data in aggregate, but there’s also an opportunity for us to create a more personal connection between drivers and their vehicles.”

He envisioned a future where the technology in the vehicle helps to create an almost palpable relationship between the car and driver.

“The technology story on the Evos [Ford’s concept car debuted at IFA this year] is really all about a car that’s connected to the cloud,” said Mascarenas. “By connecting the car to the cloud we can use all kinds of data that to make a car that really adapts to your personal lifestyle. This includes data ranging from where you like to drive to what your driving style is, all the way down to your musical preferences and situation awareness of traffic and environment.”

In essence, what Mascarenas described to me was the contextual equivalent to what Siri does, but the difference being the function that same data served. Siri serves as a personal assistant, whereas the data collected from many of the same data sources serves to create a more comfortable and personal driving experience, and communicating the information non-verbally.

Of course, most of this new-fangled technology is available only in the Evos concept car, but Cali Lewis reported from IFA this year that some of this technology will be implemented in production vehicles as soon as January 2012.

.

Does This Mean Ford’s Going After Apple?

I’d doubt it seriously, though if were cut of the same jib as some of my peers in the blogosphere, I might set this up as “Ford’s gunning for Apple” for the extra pageviews. The truth is that Ford seems to have some very prescient technologists in their employ who spotted the trend of “Cloud Convergence” far earlier than any of their competitors. Simply because Ford and Apple are the only two organizations trying to create consumer facing machine intelligence in their respective products.

Mascarenas did tell me though that competing feature for feature with Siri or any other smartphone entrant isn’t in the cards.

“It’s a safety issue,” he told me. “When you’re driving, often with other passengers and a myriad of other distractions, we must keep interactions with our software as safety oriented and transparent as possible.”

Will Ford be the last entrant in this market? I think not. Certainly Microsoft will evolve TellMe to the point where it may have an identifiable personality trait, but I don’t expect to see it on the phone. The place where TellMe is most used is on the XBox Kinect, a market that is certain to only solidify in the coming weeks and months as the new Metro interface rolls out with live TV built in; a market that Apple can’t begin to hope to put a dent in.


A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU