The "TechWeb report" of the IPhone 4S makes voice-aid Siri the new copycat target for domestic entrepreneurs. In this respect, Innovation Factory founder Lee Kai-Fu said that the domestic entrepreneurs do not encourage similar "China Siri" or "Android Siri" business plan, because the investment of high risk technology difficult to achieve, and difficult to effectively "guide user behavior."
Siri is a voice control feature Apple has used on its iphone 4S to turn the iphone 4S into an intelligent robot that uses Siri to read SMS to the restaurant to ask for weather voice setting alarms. Siri currently supports speech recognition in English, French and German, where the English part only supports British and Australian accents, and English accents in other countries and regions are difficult to identify.
The introduction of Siri features the voice control technology as a new direction of product development at home and abroad, competitors are also competing to launch similar services. Google has reportedly applied for a patent that suggests it is developing voice-control technologies similar to Apple's Siri, allowing users to control TVs via Android phones.
At the same time, the boom has also scraped into China Mobile's Internet network. There are a number of entrepreneurs trying to develop a product like voice control, even trying to replicate the model to the more open Android market.
But as a former Apple employee, Lee has suggested to such entrepreneurs in his microblog that plans like "China Siri" or "Android Siri" are less suited to start-ups because of the difficulty of investing in high risk technology.
Lee said, "A lot of recent" China Siri "" Android Siri "entrepreneurial plan. My feedback: 1 The main interface of smartphone is hand touch, voice helper solves what real user needs and sore spot? 2 how to overcome the background of massive data learning technology threshold? Only one company in the world has this technology. 3 Application who develops? own difficult to expand, with other people's application integration poor impact experience, 4 voice server and bandwidth cost, how to overcome? "
At the same time, Kai-fu Lee also made a few analysis of Apple Siri: 1 speech recognition is accurate, but the semantic understanding is difficult; 2 The semantic understanding is not enough by natural language analysis, it needs massive corpus and feedback automatic learning; 3 Siri availability is insufficient, Apple use "flirt" to obtain corpus, after learning can increase availability; 4) Noise is still a problem, Apple relies on 4S special hardware, 5 application combination difficult, suitable for Apple closed systems, 6 users of the assistant has "human intelligence" high expectations.
As we all know, Apple has become a global brand and loyal fans, even if the iphone 4S did not make a disruptive innovation, only by using Siri's lightweight services also triggered a buying spree. This is a significant (blog, microblogging) relationship with the already well-established user behavior link in Apple's eco-chain. Lee also believes that, compared with Apple Siri, domestic entrepreneurs involved in this area, the most need to solve is not the user problem, but how to effectively guide user behavior. (Chuck)