Amazon nearly doubled its capital spending to over $32 billion in the second quarter to build data centers, expand its already massive logistics network and outfit the warehouses with robots.
The Seattle-based company expects to spend up to $100 billion this year on investments related to artificial intelligence for its Amazon Web Services cloud computing service.
"We will continue to invest more capital in chips, data centers and power to pursue this unusually large opportunity that we have in generative AI," Chief Financial Officer Brian Olsavsky said during the company’s earnings call Thursday.
Meanwhile, Amazon is expanding its same-day delivery facilities to more than 4,000 smaller cities, towns and rural communities by year’s end, and installing robotics and automation across its facilities to improve fulfillment efficiency, Amazon executives said.
The e-commerce giant has now officially deployed more than 1 million robots in warehouse facilities and fulfillment centers across the globe. By contrast, the company employs just over 1.5 million people.
The tech giant saw productivity gains across its transportation network in the second quarter, driven in part by better placement of inventory. By bringing the right products closer to customers, Amazon can make them less costly and quicker to deliver, which in turn boosts sales, executives said.
The moves helped drive record sales and new memberships at Amazon’s biggest-ever Prime Day promotion in July.
“When we optimize inventory location, we can consolidate more items for package, reducing packaging materials and costs,” Olsavsky said.
Amazon’s new generative AI model called DeepFleet will coordinate the movement of robots, making its entire robot fleet smarter and more efficient. DeepFleet improves “travel time” across warehouses by 10% and allowing faster delivery to customers at lower cost.
“At our scale, it's a big deal,” CEO Andy Jassy told investors. "DeepFleet acts like a traffic management system to coordinate robots' movements to find optimal paths and reduce bottlenecks.”