Internal bureaucratism, AI is lackluster! What happened to Amazon AWS?

As the inventor and long-time absolute leader in cloud computing, Amazon is showing unprecedented fatigue. The profit engine of this Amazon company is now struggling due to the growing internal bureaucracy, slow responses in the critical artificial intelligence race, and the strong pursuit of competitors, putting its market dominance to a severe test.

The latest developments have further intensified market concerns. On Thursday, Google, a subsidiary of Alphabet, announced that it would provide up to 1 million of its dedicated AI chips to the AI startup Anthropic PBC. This move not only deepens Google's partnership with this rapidly growing AI newcomer but also poses a direct blow to Amazon, which had previously invested billions of dollars in Anthropic.

Meanwhile, the operational stability of AWS has also raised red flags. Just this Monday, the cloud division experienced one of its most severe outages in history, causing a critical data center cluster to go down, disrupting operations for hundreds of companies and consumer applications. The incident lasted about 15 hours before services were fully restored.

These events highlight deeper issues facing AWS. Nearly three years after OpenAI's ChatGPT sparked an AI frenzy, the market generally believes that AWS's performance in the AI field has lagged behind its tech peers. Although AWS remains the leader in the cloud market, its growth has shown signs of fatigue, while competitors are closing in, eating away at its once-stable territory, bringing new uncertainties to investors and the market.

Slow Start in the AI Race, Missing the Opportunity

In determining the future battlefield of AI, AWS's start has clearly lagged behind. A symbolic moment occurred on November 30, 2022, when then-AWS CEO Adam Selipsky delivered a two-hour speech at the annual re:Invent conference, but hardly mentioned AI. That afternoon, OpenAI released ChatGPT, completely disrupting the entire industry.

According to insiders, AWS was not blind to the potential of AI. It even provided strong computing power support to the startup Anthropic in its early days, but due to doubts about the profitability model of emerging technologies and a culture of reluctance to pay for external technologies, Amazon initially missed the opportunity to invest in Anthropic. This hesitation allowed competitor Google to take advantage.

It wasn't until September 2023 that Amazon made its first substantial investment in Anthropic, planning a total investment of up to $4 billion in exchange for Anthropic's commitment to use AWS's computing power and self-developed chips. To many seasoned employees who understand Amazon's aversion to paying high technology premiums, this move seems more like a "desperate measure." The deepened collaboration between Google and Anthropic on Thursday further highlights Amazon's passive position in this competition

The "Big Company Disease" of a Vast Empire

Amazon, once proud of its "startup" spirit, is now being weighed down by the bureaucracy of a vast empire. According to interviews with 23 current and former AWS employees, they generally describe an internal bureaucratic system that is slow to act when agility is needed.

A sales engineer recalled that before the pandemic, he was only six management levels away from then-CEO Jeff Bezos, but after a promotion, he found this year that he was now 15 levels away from current CEO Andy Jassy (Amazon claims this experience is an exception). In an increasingly bureaucratic culture, decision-making speed has significantly slowed. Three employees working on different AI projects stated that they were required to repeatedly revise project proposals until market conditions changed, rendering their ideas outdated.

Meanwhile, the large-scale hiring during the pandemic led to a surge in management levels, while subsequent layoffs and cost-cutting made promotions and salary increases difficult. In an industry-wide talent war for AI, AWS has lost several key executives, including heads of AI, startup sales, chip design, and data center infrastructure.

Changing Competitive Landscape, Startups "Defect"

As AWS struggles against internal "bulkiness," the external competitive landscape has dramatically changed. Microsoft's cloud business has seen a faster growth in backlog orders than Amazon. Oracle, once seen as a "runner-up" in the cloud market, has signed billion-dollar contracts to provide hosting services for cutting-edge AI development. Google has become a more threatening competitor, especially in attracting AI startups.

"If you haven't won over those native AI startups that will scale five to ten times in the future, that could be a real problem for the business," said Josh Beck, an analyst at investment bank Raymond James. Pete Schwab, who worked at Amazon for ten years, chose Google when founding his own AI company, believing that Google "does a better job of supporting small companies like ours."

Even AWS's existing major clients are "shopping around" in the AI field. Grammarly Inc., which has long operated on AWS, saw its Chief Technology Officer Mark Schaaf state in an interview that AWS's AI model marketplace Bedrock failed to meet the company's needs in terms of pricing, leading them to turn to models from OpenAI and Meta.

New Leadership at the Helm, Striving to Regain Initiative

Facing internal and external challenges, AWS is actively seeking change. This year, veteran Matt Garman, who has worked at AWS for many years, succeeded Selipsky as the new CEO. Colleagues describe him as shrewd and capable, and external analysts believe he is a more suitable leader for AWS's current "wartime" state.

After taking office, Garman urged employees to focus on delivering promised products. To speed up the process, some AI teams even set aside typical product development guidelines, sacrificing documentation and regular reviews. Over the past two years, AWS has terminated or halted the development of about thirty services and major features to free up engineers for building AI tools AWS is pinning its hopes for a comeback on AI services like Bedrock and its self-developed Trainium2 chip. Company spokesperson Selena Shen stated that Bedrock has tens of thousands of customers and is widely regarded as the company's most successful AI product. Amazon CEO Andy Jassy has also repeatedly emphasized that the cost-effectiveness of the Trainium2 chip is its potential competitive advantage, with the cost of handling AI workloads being 30% to 40% lower than that of competitors' hardware.

However, whether it's the $11 billion data center built for Anthropic in Indiana or the accelerated iteration of internal product lines, AWS must face a harsh reality: competitors that it once easily left behind have now dragged it into a fierce struggle. Google's latest collaboration with Anthropic once again indicates that Amazon has no room for an easy victory in this AI race that concerns the future