It was only a matter of time before Amazon ramped up the race to build a better voice-driven artificial intelligence (AI) experience versus a field of serious competitors, and the eCommerce titan has come out swinging with new products, and bigger plans in the pipeline.
The news comes via a series of announcements and other signals. On Wednesday (May 17) Amazon unveiled four new devices in its Alexa-powered Echo line of smart products — the Echo Pop, Echo Show 5, Echo Show 5 Kids, and Echo Buds — touted as “all-new” in a blog post.
“Customers around the world love Alexa as their trusted, personal AI— they’ve now purchased well over half a billion Alexa-enabled devices, and use of Alexa increased 35% last year,” said Rohit Prasad, senior vice president and head scientist for Alexa.
“These new devices give customers more options and more utility at incredible value. And as with every Echo device we’ve shipped, they will only continue to get better as we add even more generative AI-powered experiences for Alexa throughout the year,” Prasad said.
On the same day, Amazon senior vice president of devices and services, Dave Limp, told Bloomberg that the Alexa digital assistant “has been using a home-built set of large language models — the foundational networks that enable ChatGPT and rival technologies — to help summarize text gathered from the web and make Alexa more conversant in various languages,” saying that new ChatGPT-type features will “roll out incrementally,” and adding, “It’s not years away, but there are some things that we have to solve.”
That puts Amazon squarely in the generative AI contest against OpenAI, Microsoft, Google, and others at a time when large language models have overtaken commercial AI efforts. With developments happening fast, Alexa is upping its game to be more conversational.
As PYMNTS’ Karen Webster wrote in April, “In 2023, commerce is on the cusp of another digital commerce breakthrough — one that will take us back to the future in many ways. One where voice is the enabler of routine interactions between people and businesses and commerce transactions, in real time and securely.”
On the heels of that column came the PYMNTS report “How Consumers Want to Live in the Voice Economy,” for which we surveyed close to 3,000 U.S. consumers on their beliefs and behaviors when it comes to voice controls and voice commerce.
In that study, 58% of consumers said they would use voice because it is easier and more convenient, while 54% would use it because it is faster than typing or using a touchscreen, with 20% saying they would prefer voice for its perceived superior security. Moreover, we found that in the prior 12 months, 65% of all consumers had used voice technology.
“On average, Americans have performed six out of 35 different tasks we tested using voice technology in the last 12 months,” the study states. “Millennials have completed 11, nearly twice as many as the average and four times more than baby boomers and seniors.”
The predominant use case per PYMNTS research is talking to a smartphone voice assistant, at 38.9% of consumers. Untold millions of those voice commands undoubtedly contained phrases like “search Google” for products, places, and a plethora of other possibilities.
That point brings us back to Amazon and Alexa. According to a Monday (May 15) article, Bloomberg observed that Amazon’s job board has been running ads seeking engineers to work on “reimagining Amazon Search with an interactive conversational experience” with tempting verbiage like, “This will be a once in a generation transformation for Search.”
In Amazon’s case, it wants to use generative AI to improve product search on its marketplace, but who’s to say, with Microsoft increasing its stake in ChatGPT and GPT-4 creator OpenAI early this year in what is widely seen as a runup to giving Google a run for its money in web search.
Google has dominated web search since the early aughts, having pushed out Microsoft, Yahoo, and several others that no longer exist. Google overtook web search by zeroing in on the relevance of results, endlessly crawling and indexing the web. Voice-based LLMs like GPT-4 and coming iterations might loosen Google’s hold on search, letting Bing — and maybe even Amazon — take a bigger share.