Amazon Web Services trains its spotlight on artificial intelligence: Page 2 of 2
By Edwin Yapp December 2, 2016
Still a cloud leader
The Seattle, Washington-based firm has in recent years clearly burnished its credentials as the go-to player in the public cloud computing game, having led the idea of renting compute, networking and storage power to enterprises that need it rather than having them build and operate their own data centres.
Gartner once again rated AWS as the leader in its annual Magic Quadrant for Infrastructure-as-a-Service (IaaS), an industry rating scorecard.
“AWS has a diverse customer base and the broadest range of use cases, including enterprise and mission-critical applications. It has the largest share of compute capacity in use by paying customers – many times the aggregate size of all other providers in the market,” the research firm noted.
AWS has certainly come a long way since its inception. But while many believed it was the go-to provider for startups, some industry observers doubted whether it could serve enterprise effectively, given that its revenue was mainly led by startups.
It also didn’t help that AWS wasn’t publically breaking out its earnings as a division of its parent Amazon.com. But that all changed in 2015, when it began doing so, not only surpassing analysts’ and investors’ expectation but thriving at raking in positive operating income results.
Its latest earning results showed a continued upward march with the company recording an US$11.1 billion in revenue during the 12-month period ending in September 2016. This puts the company surpassing its publicly stated goal of having a revenue run-rate of close to US$13 billion in its fiscal year.
Revenue from its third quarter alone amounted to $3.23 billion, growing by 55% year over year. Last quarter’s operating profit stood at US$861 million, growing by 101% year over year, which represented an operating profit margin of 27%.
Defending its turf?
But as impressive as the last few years has been for AWS, the challenge is not only to continue growing revenue through its broad array of cloud services but also how to build on these basic cloud services.
This is where the next battleground is taking place, as old stalwarts the likes of IBM, Microsoft Corp, Google Inc is heating up in the artificial intelligence (AI) and machine learning space.
All three giants are investing heavily in this nascent field, with arguably IBM having a marginal head start these past few years.
In 2011, Big Blue showcased its AI prowess by pitting its IBM Watson AI computer against two of Jeopardy’s greatest winners. In that encounter, IBM Watson defeated them, staking its claim as the winner of general knowledge, query-based game show. Meanwhile, Google also did the same this year, when its AI machine called AlphaGo defeated Lee Sedol in a highly complex decision-making game known as Go.
The introduction of these new AI services is seen by some industry observers as a move by AWS to not only defend its turf in the cloud arena but also to show itself to be a credible player in the nascent field of AI and machine learning.
Some industry watchers have also argued that the aforementioned competitors are also perceived to be ahead of AWS in this field. The announcements made by Jassy were in part to dispel this notion.
According to Naveen Chhabra, senior analyst at Forrester Research, the speed of innovation and the introduction of such new AI services, which are market leading and disrupting, augers well for AWS.
Speaking to Digital News Asia (DNA) on the sidelines of re:Invent 2016, the analyst said, “The use of the Alexa AI could be used to develop enterprise application such as in an IT service environment, where new services can be built and deployed in days rather than in weeks.”
Naveen said the beauty of using such an AI service is that it eliminates the need for human intervention and leverages automation, something that will increase process efficiencies, he noted.
“This is the next step for cloud applications,” he argued.
Asked if he thought this was a defensive move by AWS, Naveen said that it was highly possible, noting that Google already has such services including image recognition and voice assistance, and that these three new services introduced by AWS compete with what Google has.
Despite this, Naveen was quick to add that this doesn’t mean that Google has the upper hand in this nascent AI space.
“Google’s AI-based services are known more in the consumer realm but its recognition is slow in the enterprise space,” he argued. “In Google’s case, the bundling of those individual consumer service as an enterprise service is yet to be seen,” he claimed.
Is AI useful yet?
Still there are others who believe that AI-based technologies in the enterprise space are still too new to be useful.
Judith Hurwitz, an analyst with Hurwitz & Associates, is one such person who isn't convinced about the importance of AI services in the cloud at this point.
Speaking to Computerworld, she argued, “I am not sure that any of these [cloud] companies are doing anything truly meaningful with AI. There is increment tooling being added to Google, for example, but this isn't transformational yet.
“Microsoft is still quite early in its ability to use machine learning and AI in its solutions. IBM is probably further ahead with its Watson platform, but we are still early in adopting AI and machine learning.”
Countering this point was Patrick Moorhead of Moor Insights & Strategy, who said that the moves AWS is taking are important first steps in the world of AI.
“Improvements to work with enterprise workloads are very important," he argued. "Forward thinking enterprises are asking for these services. Insurance companies want these to categorise photos of accidents. Pizza chains want you to be able to order a pizza through your favourite messenger in plain language."
“Amazon actually isn't ahead of their cloud competitors in this area," said Moorhead. “They're behind. Microsoft Azure has many more services available and has had them for a longer period of time.”
Edwin Yapp reports from AWS re:Invent 2016 in Las Vegas, at the invitation of Amazon.com Inc. All editorials are independent.