Amazon's earnings suggest that the shift away from training towards inference and agents means their bet on Trainium is paying off. Plus, additional notes on ads, agents, and sports rights.
Here's something that might surprise you — Amazon's latest earnings reveal a clear shift. According to Ben Thompson, writing in TechCrunch, their focus is moving away from training AI models toward inference and deploying agents. And this pivot seems to be paying off, with Trainium — Amazon’s custom chip — seeing stronger demand. The reason? Companies are less interested in training massive models and more eager to run and use them in real time. Thompson points out that this shift isn’t just technical; it’s strategic, making Amazon’s infrastructure more efficient and profitable. Plus, there’s more happening behind the scenes — Amazon’s ad business continues to grow, and they’re even making plays in sports rights and AI agents. But here’s the thing — this isn’t just about hardware or ads. It’s a sign that the AI race is changing. The winners now are those who can turn models into everyday tools, fast and flexible. The takeaway? In the AI game, moving quickly from development to deployment is everything.
Amazon's earnings suggest that the shift away from training towards inference and agents means their bet on Trainium is paying off. Plus, additional notes on ads, agents, and sports rights.
Amazon's earnings suggest that the shift away from training towards inference and agents means their bet on Trainium is paying off. Plus, additional notes on ads, agents, and sports rights.