亚马逊AGI组织裁员，AWS的AI估值锚要换了吗？

TL;DR

• Amazon confirms the elimination of some positions within its AGI organization, while stating that large AI models remain a key focus.
• The market is divided on whether this adjustment is a downgrade of its in-house model efforts or a refocus towards customer-funded projects.
• Associated tickers: AMZN, AWS, Anthropic, AI cloud infrastructure chain.

On July 22, Amazon confirmed the elimination of some positions within its AGI organization, while stating that the company is still building large AI models, calling it one of its most important tasks. According to a report cited by Reuters, Amazon explained the adjustment as refocusing resources on the areas most critical to its customers' future.

The specific number of layoffs was not disclosed, nor can it be directly interpreted as Amazon abandoning AGI. It more accurately brings the contradictions within Amazon's AI narrative to the forefront: Big Tech is still investing hundreds of billions of dollars in AI infrastructure, but the teams closest to the long-term AI ambition are beginning to experience organizational contraction.

For investors, the question isn't how many people Amazon laid off, but that AWS's AI valuation anchor is being reassessed. Previously, the market was willing to pay a premium for Big Tech's AI investments under the assumption that stronger model capabilities would lead to greater future revenue. Now, the more pressing question is when these investments will translate into customer payments, cloud revenue, and improved profit margins.

AGI (Artificial General Intelligence) can be simply understood as a yet-unrealized long-term goal where AI can learn and solve problems across different domains like a human. It represents long-term vision, but doesn't necessarily translate into immediate revenue. What AWS needs is to package AI capabilities into services that enterprises can buy, use, and customize right now.

Greater AI Investment, Harder Organizational Choices

The key to this layoff isn't whether Amazon will continue developing models. The official statement has already drawn the boundary: large models remain a priority, but resources will be allocated to projects with the highest customer focus and priority.

Organizational actions indicate that while AI investment hasn't stopped, the margin for error is shrinking. In December 2025, Amazon adjusted its AI leadership, with Andy Jassy announcing that Peter DeSantis would lead a new organization covering AI models, chips, and quantum computing. Rohit Prasad left in late 2025, and Pieter Abbeel took over frontier model research within AGI. A report cited by Reuters also noted that AGI Lab head David Luan left in February 2026.

Viewed together, these changes show Amazon isn't exiting the AI arms race, but is reordering its internal investment portfolio. Long-term research retains its narrative value, but projects closer to customers, revenue, and productization are gaining higher priority.

This is also the common backdrop for Big Tech AI trades. Over the past two years, the market primarily traded on who dared to spend, who had computing power, and who had the model. Now, capital expenditure alone is no longer sufficiently scarce; investors are starting to ask about returns: Can model teams, chips, data centers, and talent ultimately crystallize into revenue?

Nova Forge Provides a Commercialization Anchor

To understand this adjustment, one needs to look at Nova Forge, released by AWS during the AWS re:Invent event in December 2025. It's not an ordinary chatbot, but a service designed to help enterprises train custom models.

In the traditional path, if an enterprise wanted a frontier model suitable for its industry, the options were either to train from scratch (extremely costly) or fine-tune an existing model (limited capability and controllability). Nova Forge's approach allows clients to start from checkpoints (intermediate training archives) within the Amazon Nova model training process, integrating their own data and Amazon's curated datasets at different training stages.

Amazon calls this open training. Simply put, enterprises don't need to build a large model from zero; instead, they start from a model base that Amazon has already trained to a certain stage and inject their industry knowledge early. This allows them to inherit foundational capabilities while becoming more specialized in their domain.

This path is crucial for AWS because it attempts to turn model capabilities into cloud service products. Clients aren't just calling a model API; they are training, hosting, deploying, and optimizing their own models on AWS. If the product works, it can drive computing consumption, platform stickiness, and subsequent operational revenue.

However, available information doesn't prove that the resources from the eliminated AGI positions have shifted to Nova Forge. A safer judgment is that the AGI organizational adjustment coincides with the emergence of customer-oriented products like Nova Forge, indicating Amazon's tendency to increase the weight of commercialization projects.

AWS's Competitive Focus Shifts to Customer Customization

Amazon's position in the foundation model race has always been unique. It develops its own Nova model, invests in Anthropic, and must maintain AWS's neutrality and model ecosystem as a cloud platform.

This dictates that AWS doesn't necessarily need to win solely with the world's best model. For enterprise clients, model leaderboards are important but not the only criterion. More practical questions are: Can it access internal enterprise data? Can it meet security and compliance requirements? Can it reduce training costs? Can it operate alongside existing cloud services?

Nova Forge precisely addresses this competitive logic. It shifts the battleground from generic model capability rankings to enabling enterprises to train their own models at lower costs. If this path succeeds, AWS can embed AI revenue into its core cloud computing business, rather than solely betting on a consumer-grade AI product.

This also explains why Amazon is maintaining its AGI narrative while simultaneously cutting some positions. The former retains long-term technological vision, while the latter forces teams to allocate resources towards directions more easily validated by customer demand.

For AMZN, the market will ultimately not just look at whether Amazon has an AGI team. More importantly, it needs to see whether AWS can prove that its AI services boost customer spending, enhance stickiness, without significantly dragging down profit margins.

Orders and Profit Margins Will Provide the Answer

This layoff is easily framed as one of two extremes: either it's Amazon's AI failure, or it's an insignificant routine optimization. The available information supports neither conclusion.

A more reasonable judgment is that Amazon remains in the AI arms race, but internal budgets and talent allocation are shifting towards directions that can be sold to customers. This change is meaningful for investors because AMZN's AI premium will increasingly depend on AWS's commercialization results, rather than purely on model narrative.

Verification points will come down to specific matters. Can Nova Forge secure real enterprise clients? Are clients willing to pay continuously? Do the trained models offer better value than standard fine-tuning? These factors will determine whether it is an effective product.

Another variable is talent attrition. If the AGI organizational adjustment merely optimizes non-critical roles, the impact is limited. If core research and engineering talent leaves, Amazon's long-term competitiveness in developing its own models will be weakened. The tension between official messaging and organizational reality will ultimately be resolved through product adoption rates, AWS AI revenue, and the return on capital expenditure in subsequent financial reports.