updated 12:50 pm EDT, Mon October 3, 2011
Siri co-founder claims tech uses 'real' AI
The Assistant voice command feature in iOS 5 will be a "world-changing event," according to Siri co-founder Norman Winarsky. Siri's technology is believed to be core to Assistant, and why Apple bought the developer some time ago. "Make no mistake: Apple's 'mainstreaming' artificial intelligence in the form of a Virtual Personal Assistant is a groundbreaking event," says Winarsky in an interview. "I'd go so far as to say it is a world-changing event."
"Right now a few people dabble in partial AI-enabled apps like Google Voice Actions, Vlingo or Nuance Go," he adds. "Siri was many iterations ahead of these technologies, or at least it was two years ago. This is REAL AI with REAL market use. If the rumors are true, Apple will enable millions upon millions of people to interact with machines with natural language. The PAL [Personal Assistant that Learns] will get things done and this is only the tip of the iceberg. We're talking another technology revolution. A new computing paradigm shift."
While Winarsky is not working at Apple, he suggests that the Siri AI may explain why Apple waited until the fall to launch a new iPhone, instead of sticking to its normal summer schedule. "I'm not familiar with Apple's roadmap and any delays but I can say that AI takes a lot of computing power," he explains. "The Siri software needs to cache data, needs to access a big dataset at wide bandwidth and needs a big processor to crunch all of the numbers. When we originally released Siri for the iPhone 3GS, we had to perform all kinds of optimizations and shortcuts to get it to work efficiently. All I can say is that it will likely run much better on a faster phone." Assistant has been rumored as only working on the next-generation iPhone because of RAM and processor requirements.
Also mentioned in the interview is that Apple finalized the buyout of Siri only two months after its signature app went public. "You can probably draw your own conclusions from that," Winarsky comments. He also observes that the speech recognition technology Apple has licensed from Nuance is likely "a lot less important than you'd probably think," because the recognition component in Siri's technology is modular. "Theoretically, if a better speech recognition comes along (or Apple buys one), they [Apple] could likely replace Nuance without too much trouble. That being said, Nuance has far and away the most IP in speech synthesis technologies in the industry. We should know, SRI launched Nuance as one of our incubated companies in 1995 and it IPO'd in 2000." Winarsky is currently with SRI, the company that originally spawned Siri, taking advantage of its coordinating role in the US government's CALO (Cognitive Assistant that Learns and Organizes) program. CALO ended in 2007.