Amazon.com (AMZN.US), the US e-commerce and cloud computing leader, is reducing staff within its strategically important robotics division. Some Wall Street analysts interpret this move, combined with Amazon's recent announcement of large-scale trials for its self-developed AI chips—specifically the Trainium and Inferentia AI ASIC compute cluster infrastructure to develop and iteratively update its proprietary large AI models—as a significant signal that the e-commerce and cloud giant is advancing broader cost-cutting measures and shifting its expenditure focus entirely toward AI computing infrastructure. Concurrently, Amazon is increasingly relying on automation systems to support its fulfillment network.
According to media reports citing informed sources, this week's layoffs affected "certain robotics roles," but the company continues to actively hire and invest in "multiple strategic areas." These latest job cuts, which bring the total number of corporate positions eliminated at Amazon since 2022 to 57,000, coincide with the company significantly increasing its large-scale investments in artificial intelligence, data centers, and humanoid robotics to maintain its crucial position in the AI race and the broader physical AI trend.
Amazon is initiating an AI cost revolution, striving to gain autonomy over training and inference costs. This action does not indicate a de-prioritization of robotics initiatives but rather a scaling back of certain robotics projects or roles with longer return cycles. Resources are being reallocated toward AWS cloud resources, AI data centers, and the self-developed AI ASIC chip ecosystem. Amazon aims for "model and chip co-design" to control its own training and inference cost structures, rather than being perpetually influenced by external GPU pricing.
Undoubtedly, as Anthropic, often termed an "OpenAI rival," plans to spend hundreds of billions of dollars to acquire 1 million TPU chips, and Meta Platforms considers a multi-billion dollar purchase of Google's TPU AI compute infrastructure for its massive AI data center build-out by late 2026 or 2027, alongside Amazon's announcement to utilize Trainium and Inferentia for developing large AI models, these developments collectively illustrate that as cloud giants launch an "AI compute cost revolution" to advance the adoption scale of AI ASICs, market concerns regarding Nvidia's growth prospects are valid.
While reducing a relatively small number of roles in its robotics team, the company is directing its 2026 capital expenditure toward approximately $200 billion, primarily focused on AWS's core cloud systems and massive AI workloads. Simultaneously, AWS continues to advance self-developed AI compute like Trainium and Inferentia, and Amazon's operations network has already deployed over 1 million robots, utilizing generative AI models like DeepFleet to enhance robot scheduling efficiency.
During the company's recent earnings call, Amazon CEO Andy Jassy confirmed an investment of roughly $200 billion. This capital will be allocated across the company but primarily toward Amazon Web Services (the AWS cloud division) because "our capacity requirements are very significant, customers really want AWS to run their core workloads and massive AI task workloads, and we can monetize the capacity we install as quickly as we can install it at scale." Jassy also described the robotics business as "a big project" for the company. With over 1 million robots in its fulfillment logistics network, automation will handle repetitive and potentially hazardous tasks to significantly boost productivity and efficiency. "We will continue to optimize inventory placement to reduce transportation distances, decrease the number of handlings per package, and significantly improve package consolidation, while introducing more advanced robotics and automation technology to increase efficiency and enhance the customer experience," Jassy stated on the earnings call.
However, this decision to scale back the robotics division comes just weeks after Amazon discontinued development of its multi-armed robot portfolio, "Blue Jay," which was initially expected to be widely deployed in Amazon's same-day delivery warehouses.
AI compute infrastructure is now the top priority. Amazon's management is effectively reallocating capital and talent from robotics projects with longer payback periods and complex engineering integration toward the AI compute infrastructure layer, which promises faster monetization. Amazon's confirmation of layoffs in the robotics department this week follows another round of significant job cuts in January. At the same time, Amazon has raised its 2026 capital expenditure target to $200 billion, explicitly earmarked primarily for AWS and AI compute infrastructure.
On the other hand, Amazon has not abandoned its warehousing automation ambitions: the company officially announced last year that its operations network had deployed its 1 millionth robot and introduced the generative AI model DeepFleet for scheduling robot fleets, claiming it can improve fleet movement efficiency by 10%. This suggests the cuts target robotics projects or roles with insufficient marginal returns rather than the "automation strategy" itself.
In other words, Amazon's current cost plan resembles a typical tech stack realignment: prioritize building a universal AI platform and a self-developed compute foundation first, then use this "cheap and scalable intelligence" to enhance robotics and the fulfillment network. This is not "robotics losing to AI," but rather robotics being integrated as a downstream application layer of an AI-platform strategy. Examining the fundamental relationship between robotics and AI data centers, Amazon appears to be acknowledging a reality: the future core bottleneck lies first in compute economics, and second in end-point automation forms. Robotics remains important, but within Amazon's ecosystem, it increasingly functions as a downstream execution layer. What truly determines scaling speed, unit cost, and iteration efficiency is the upstream ability to train/deploy models at lower cost and replicate these capabilities for AWS customers, Nova, Alexa, Rufus, as well as warehouse scheduling and robot control.
Amazon's stock price rose nearly 4% by the close of trading on Wednesday, marking its best single-day performance since November. This gain was primarily fueled by a technical rebound in oversold tech stocks amid improved market risk appetite, coupled with US service sector growth hitting its fastest pace since mid-2022 while price pressures eased, and better-than-expected ADP employment data. These strong economic indicators temporarily overshadowed macroeconomic concerns stemming from Middle East geopolitical tensions. The three major US stock indices all closed higher, while US Treasuries and the US dollar fell; another risk asset, cryptocurrencies, also surged accordingly.
Comments