🤖 Amazon CEO Defends $200 Billion AI Spending Spree

💸 Amazon CEO Defends $200 Billion AI Spending Spree

🚖 Are Self-Driving Cars Safe? Waymo’s NEW Safety Stats.

📍 6 AI Stock Sectors

🛠️ 5 Trending Tools

🥪 4 Brief Bites

💰 3 Funding Frontlines

💼 4 Job Opportunities

While investors have become somewhat reluctant to buy into his promises of future AI-generated cash flows, Jassy pushed back on those reservations, claiming Amazon isn’t investing “on a hunch.” He provided a basic timeframe for when the $200 billion AI spending spree will pay off.

Much of this CapEx is expected to be monetized in 2027 and 2028, and is already supported by customer commitments, including more than $100 billion from OpenAI, as well as several other strategic partnerships “deep in process.”

For context, Amazon invested $50 billion in OpenAI in exchange for its adoption of AWS’s “Trainium3,” a purpose-built AI accelerator fabricated on TSMC’s 3nm process technology, which offers higher transistor density to deliver more computing power with faster processing speeds.

AWS, which accounted for nearly 60% of Amazon’s operating profit last year, still relies on NVIDIA’s cutting-edge GPUs to power over 200 fully featured cloud services. “Virtually all AI thus far has been done on NVIDIA’s GPUs, but a new shift has started,” said Jassy. “AWS customers want better price-performance.”

Amazon expects its AWS Trainium Family to save the online retail giant tens of billions of CapEx dollars per year, along with several hundred basis points of operating margin advantage. AWS’s custom chip business achieved an annual run rate revenue above $20 billion and is growing in “triple-digit percentages.”

Waymo relies on a multi-sensor suite:

This multi-sensor suite constantly captures real-world data that’s interpreted by AI-powered algorithms and compared to detailed 3D maps of surrounding areas, enabling the Jaguar I-PACE SUVs to autonomously steer, brake, signal, accelerate, and anticipate future actions (e.g., “That pedestrian’s about to jaywalk!”).

When analyzing Waymo’s performance across 170.7M rider-only miles compared to human-driven vehicles over the same distance in the same cities, it achieved 92% fewer pedestrian crashes with injuries, 85% fewer cyclist crashes with injuries, and 83% fewer airbag deployments.

The verdict seems clear: self-driving cars are safer than human-driven vehicles, right? This data comes with an important caveat: the sample size remains limited. Since fatal crashes are fortunately rare, more miles are needed to gain greater confidence. While Waymo has driven nearly 200 million fully autonomous miles, much of that mileage took place in controlled urban settings within pre-mapped geographic areas.

We must agree on a threshold of acceptable harm for self-driving cars. For instance, we’re willing to tolerate human error, creating cultural norms that accept imperfections, like saying: “No one’s perfect!” But will we hold self-driving cars to the same imperfect standard? In other words, self-driving cars are still SUPER promising, but they’re not perfect, just better. We must get comfortbale viewing them as “low risk,” not “zero risk.”

Pomo raised a $4.5M Seed Round to monitor your market 24×7.

Patlytics closed a $40M Series B for AI-powered patent portfolios.

Modus landed an $85M Series A for AI-native accounting audits.

GoodRx (San Francisco, CA): Data Engineer Intern, Summer 2026

NVIDIA (Westford, MA): ASIC Design Engineer, NVLink, Entry-Level

OpenAI (San Francisco, CA): AI Support Engineer, GTM, Mid-Level

Harvey (Dallas, TX): Sr. Payroll and Equity Analyst, Senior-Level