Last week, multiple stories stated that Amazon’s Alexa was a failure and the company could lose as much as $10 billion on this technology this year.
The Ars Technica article quotes a Business Insider report that says, “The BI report spoke with “a dozen current and former employees on the company’s hardware team,” who described “a division in crisis.” Just about every plan to monetize Alexa has failed, with one former employee calling Alexa “a colossal failure of imagination,” and “a wasted opportunity.” This month’s layoffs are the end result of years of trying to turn things around. Alexa was given a huge runway at the company, back when it was reportedly the “pet project” of former CEO Jeff Bezos. An all-hands crisis meeting took place in 2019 to try to turn the monetization problem around, but that was fruitless. By late 2019, Alexa saw a hiring freeze, and Bezos started to lose interest in the project around 2020. Of course, Amazon now has an entirely new CEO, Andy Jassy, who apparently isn’t as interested in protecting Alexa.”
I have followed the voice recognition concept since 1992 when I was asked to go to Apple and meet with a young Chinese engineer doing pioneering work on this technology. His name is Kai-Fu Lee, who was eventually stolen from Apple by Microsoft to work on a similar project, and then Google poached him to start Google’s China business which he ran for three years before going out on his own.
The reason Apple brought Mr. Lee to work on voice technology was for the purpose of adding it to the Mac and eventually the Newton. The project was called Casper and first demos were on a Macintosh. However their Newton Message Pad was also in its sights.
When Apple introduced the Newton it was pen-based, but then Apple CEO, John Sculley wanted to enhance Newton’s user interface with voice too.
Newton as a pen-based handheld failed miserably. Ultimately Casper was scuttled because the truth of the matter is that in the case of voice recognition, the technology was not there to make it possible to understand voice commands accurately on the Mac or even the Newton.
Although there was a lot of voice recognition research before 1992, as far as I can tell, Mr. Lee’s Apple project was one of the first PC-focused works on voice recognition for commercial use in a handheld computer.
In early 2000, after Steve Jobs returned to Apple, the company started another series of voice recognition projects. Siri’s roots can be traced back to this early work. However, its focus was on being another type of user interface for the Mac. Also, Steve Jobs was eyeing another market which was to provide tools for the blind to use a Mac as a personal computer.
Fast forward to today, and voice recognition is mainstream, especially as a user interface. As I drive, I can tell Siri in Apple Car Play to send a message, make a call or set a new driving destination. I can do the same with Google’s voice programs. With Amazon’s Alexa, I can call up my music on my Echo device, turn on lights or set a new room temperature by voice. However, in all these examples, voice as a user interface is the purpose of voice recognition.
On the other hand, using voice to buy products was mainly a nonstarter for voice from the beginning. Amazon especially tied voice to purchase goods on Amazon, but as it is now clear, that business model for Alexa never took off.
I am not surprised. I have never used Alexa to buy anything from Amazon, and I am what Amazon calls a power user as I spend thousands a year on their site. Most people do not buy products and services from Google voice commands either.
As for Apple’s Siri, Apple never pushed for Siri to be used to buy apps and services. As I stated earlier, Apple’s focus for Siri was always as a voice user interface and not to drive retail commerce. Instead, they built great products, services, an app ecosystem, and physical stores.
I am unsure why Amazon was willing to spend billions on Alexa beyond its prominent role as a user interface. But as I said earlier, I always felt that using Alexa as a means to buy products would never be fruitful and pay off for Amazon.