Amazon Bets on Cost-Effective AI Strategy with In-House Chips
Amazon's newly appointed AI chief Peter DeSantis is spearheading a low-cost approach to artificial intelligence, leveraging the company's proprietary Trainium and Inferentia chips. This strategic pivot aims to undercut competitors on price while addressing investor concerns over Amazon's $200 billion capital expenditure plan.
The move comes as Amazon's stock has declined 8% year-to-date, reflecting market unease about heavy AI infrastructure spending. DeSantis, a 28-year Amazon veteran who previously built AWS chip operations, argues that reducing AI costs is essential for widespread adoption.
While Amazon's Nova AI model has underperformed benchmarks, the company is developing Nova 2 to improve competitiveness. The departure of AGI Lab head David Luan adds another LAYER of complexity to Amazon's AI ambitions during this transitional period.