Seattle, Wash.
Matt Garman helps make your on-line life – your Starbucks order, your Netflix binge, your Pinterest shopping– potential.
As the CEO of Amazon’s cloud computing division, he performs a serious position in deciding the way to deploy the computing assets wanted to drive the web – a accountability that would quickly embody shaping the way forward for AI.
Few folks know the enterprise in addition to Garman, who sat down with NCS at Amazon’s Seattle headquarters in mid-March. As an intern, he wrote the paper defining Amazon Web Service’s enterprise technique. And he was AWS’ first product supervisor when he joined Amazon full time in 2006, serving to corporations pivot to the internet in the web’s early days.
Twenty years after its March 2006 launch, AWS has grow to be essential for nearly any firm that depends on internet-powered instruments. When AWS goes down, components of society grind to a halt.
That’s big enterprise for Amazon, raking in $128.7 billion in gross sales final 12 months. But AI has upended the tech business. Amazon’s already making main adjustments, like ratcheting up AI infrastructure spending to an anticipated $200 billion this 12 months and axing tens of 1000’s of jobs.
Garman defined why Amazon views the strikes as crucial. The downsizing is rushing up each day operations, and there’s sufficient pent-up AI demand to maintain Amazon’s cloud instruments busy for the subsequent 5 to 10 years even when the tech stops advancing, he stated.
Back when AWS launched, he stated, “we had to explain what even the concept of cloud computing was, and why it was a thing, and why it’s something that Amazon would be involved in.”
Those conversations really feel acquainted once more.
“Just fast forward to where we are today in AI, I think (it’s) actually much the same challenges,” Garman stated. “A lot of people are going to have to think about how they do work differently.”
AWS launched to offer digital IT infrastructure and servers to companies.
The pitch was easy: Amazon would deal with backend technical operations in order that corporations may give attention to their merchandise and prospects. But it was a dangerous bet for a corporation that rose to fame reshaping the retail business.
“We’re very comfortable being misunderstood,” Amazon founder Jeff Bezos informed Bloomberg Businessweek in 2006.
Amazonians nonetheless embrace that considering as we speak, Jeff Barr, chief evangelist for AWS and a 23-year veteran of the firm, informed NCS.
(*20*)
Walking by way of the reception space on my technique to meet Garman, I handed a barista serving particular lattes celebrating AWS’ twentieth birthday.
We’re greeted by a disembodied voice over an intercom asking who we’re right here to see on the restricted flooring the place Garman awaits. As I’m escorted to one among his private convention rooms, I see a bodily signal of AWS’s huge attain: a shelf adorned with autographed soccer helmets – the NFL is one among AWS’ greatest purchasers, a testomony to how essential it’s grow to be to life each on-line and off.
It’s a great distance from the conversations about web storage in a Seattle pub that gave rise to AWS. Garman tells me even his personal mother and father struggled to grasp what cloud computing was when he described his job to them again then.
“(It) was super hard to explain to them,” he stated. “And my dad’s like, ‘Is it like the guy who comes to my office and fixes the printer?’”
But AWS ultimately grew to become indispensable to internet entrepreneurs – and now its future relies on doing the similar for AI corporations.
Amazon is closely concerned with a few of AI’s greatest gamers, like OpenAI and Anthropic, investing billions straight, serving to distribute their companies and offering tech for coaching their fashions. It’s additionally developed customized chips for AI duties.
But AWS additionally desires to be important for all corporations by way of its Bedrock platform, which Amazon says greater than 100,000 corporations use to construct their very own AI apps and brokers. Just as AWS’ early merchandise allowed corporations to entry storage and computing energy with out making big infrastructure investments, Amazon’s newer instruments are making AI fashions extra accessible, the firm says.
Amazon is the largest cloud supplier, placing it in a powerful place to money in on AI-related surges in computing demand. But rivals Microsoft and Google are desirous to catch up.
Amazon’s share of the cloud market dipped from 39% in 2023 to 37.7% in 2024, based on market analysis agency Gartner.
Google’s cloud is at the moment enticing to startups as a result of it’s barely simpler to make use of and get began with, based on Jacob Colker, managing director of the Seattle-based A12 Incubator, which helps AI startups construct their companies. Plus, Google has a extra beneficiant credit score program for younger corporations.
Yet the tide may shift at any second.
“The pace of innovation, obviously, is breakneck in the world of tech, and I think that’s equally true for a lot of the cloud providers,” he stated.
Amazon’s monolithic information facilities and hundreds of thousands of miles of fiber optic cables underpin the web. But it’s the selections made inside buildings like the glassy Amazon Reinvent tower in downtown Seattle that assist decide Amazon’s future – and that of hundreds of thousands of different corporations, if issues go Amazon’s method.
Within its partitions on an overcast Seattle afternoon, Garman and I focus on one among the greatest questions Wall Street has for Amazon: the $200 billion it expects to spend on capital expenditures this 12 months associated to AI infrastructure. The quantity was greater than $50 billion increased than analysts’ expectations and the $131.8 billion Amazon spent on property and tools in 2025.
Spending on AI has grow to be so massive that market analysis agency Gartner now categorizes corporations like Amazon as “digital nation states” since they “control enough land, power, water (and) talent to actually rival countries,” stated analyst Nicole Greene.
Where is that cash going?
“This is not a secret, it’s data centers and servers,” Garman stated.
The sky-high spending has fueled issues about an AI bubble.
Tech giants insist AI demand is so feverish that they’re scrambling for compute energy. Critics need to know when corporations are going to see a return on these investments.
Amazon’s spending spree isn’t its solely eyebrow-raising transfer not too long ago. The firm slashed roughly 30,000 jobs throughout two rounds of layoffs – one in October and one other in January – to maneuver quicker as AI evolves, it claims.
Amazon previously said AI developments didn’t drive most of the cuts, though CEO Andy Jassy said in June that the firm will want “fewer people” as AI adjustments work. Garman stated AI is enjoying an even bigger position in operations, like provide chain planning, managing information middle assets and, after all, software program improvement.
AI coding instruments that enable programmers to construct their very own private workforces of AI brokers are upending the software program business. Projects that may have taken two to 3 years at AWS are actually being executed in a matter of months with small groups, Garman stated.
In truth, Garman stated AWS’ groups are actually “building at a rate that we haven’t seen for many years.”
Amazon’s strikes might have struck a nerve as a result of they hit on two of the greatest AI-driven issues to emerge over the previous 12 months: the expertise’s impression on jobs and whether or not it’s making a bubble.
But some consultants, like cofounder of the Stanford Institute for Human-Centered AI James Landay, previously told NCS the position AI is enjoying in software program improvement has in some instances been overstated. And with each earnings name, analysts appear to have extra questions on when the billions being poured into AI will present up in new merchandise.
Garman is assured these bets will repay.
At a latest assembly with round 150 senior expertise leaders, 90% of the attendees raised their palms when requested in the event that they had been both seeing a “solidly positive” return on AI investments or anticipate to inside the subsequent six months, he stated.
“I’m sure they exist,” Garman stated about indicators of an AI bubble. “But I have not seen them yet.”
-NCS’s Rhyannon Bartlett-Imadegawa contributed to this report.