WSJ Tech News Briefing

Episode: The Almost-Crisis That Struck Virginia’s Power Grid
Date: March 3, 2026
Host: Peter Ciampelli
Featured Guests: Sean McClain, Katherine Blunt
Main Theme: How data centers’ surging power demands—and their sudden drop-offs—are creating new grid risks, plus Amazon’s new efficiency-focused AI strategy.

Episode Overview

This episode of WSJ’s Tech News Briefing uncovers two pressing issues at the intersection of technology and infrastructure: Amazon’s attempt to catch up in the AI arms race with a new cost-focused leader, and exclusive reporting on a near-crisis in Virginia’s power grid triggered by the abrupt power-down of multiple data centers. The episode provides insight into both the competitive cost dynamics shaping big tech’s future and the less-visible vulnerabilities in America’s electricity infrastructure as it grapples with the rise of energy-hungry data centers.

Key Discussion Points and Insights

1. Amazon’s New AI Strategy

[00:17 – 05:02]

• Amazon Appoints Peter DeSantis as AI Chief
  • Amazon is trying to regain ground in the AI platform race by putting Peter DeSantis, a company veteran, in charge.
  • DeSantis is seen as a “celebrity figure” internally, credited with Amazon’s cloud and chip infrastructure successes. (Sean McClain, 01:32)
• Strategic Shift: Cost-Efficient, Task-Specific AI
  • DeSantis aims to combine AI, data center, and chip efforts to accelerate product development and reduce delivery costs.
  • Focus is on building AI that’s “cheaper to sell to businesses looking for specific tasks” rather than general-purpose competitors like ChatGPT. (Sean McClain, 02:29)
  • “There’s a lot of tasks that benefit from AI but don’t need the world’s strongest or world’s most cutting edge model to do well.” (Sean McClain, 02:41)
• Customer Example – Nimbus Therapeutics
  • Nimbus used Amazon’s Nova AI to design new molecules for drug development, finding it “just as good” as Anthropic’s Claude at one-tenth the price. (Sean McClain, 03:33)
• Investor Concerns Over Spending
  • Amazon is planning massive capital expenditures—$200 billion this year alone, primarily for AI infrastructure.
  • This heavy investment puts pressure on Amazon to deliver, with worries over whether customer adoption will keep up.
  • “Everybody is…price to perfection in this market. So if you’re not executing, then your stock market will reflect that instantaneously.” (Sean McClain, 04:10)

2. Virginia Data Center Near-Crisis: Risks to Power Grids

[05:50 – 11:01]

• The Incident: 42 Data Centers Suddenly Go Offline
  • In early 2025, a transmission line disturbance led to 42 data centers in Virginia switching to backup power at once—an event equal to the sudden loss of enough electricity to power a million homes. (Katherine Blunt, 06:22)
  • “If one or two do that, that's not really an issue. But to have so many go offline all at once created sort of the prospect of instability on the broader power grid.” (Katherine Blunt, 06:22)
• Why Dropping Demand is a Problem
  • Electricity supply and demand must be balanced at all times. If demand drops suddenly, it can flood the grid with excess energy, risking cascading failures and power plant damage.
  • “If there’s too much supply…and that demand is no longer there, you have the same sort of risk to the rest of the system and the infrastructure.” (Katherine Blunt, 07:14)
• Operator Response and Industry Concern
  • Grid operator PJM Interconnection managed the event without catastrophe thanks to contingency plans.
  • Concern escalates as more and more large data centers come online; a wider or more severe repeat could outstrip the grid’s ability to respond.
  • There were two similar incidents in a matter of months: July 2024 and February 2025. (Katherine Blunt, 08:12)
  • The industry is debating how to ensure data centers can “ride through” minor power disturbances instead of immediately switching off-grid. (Katherine Blunt, 09:53)
• Possible Solutions
  • Dominion Energy is working with tech companies to recalibrate data center sensors to “withstand certain disturbances… rather than immediately switching to backup.” (Katherine Blunt, 09:53)
• Potential Impact on the Public
  • If the grid operator can’t manage a sudden loss of demand, “power plants would trip offline, they would probably be subject to some sort of significant damage. And so to repair that would take a lot. That’s extreme. But that's what could potentially happen.” (Katherine Blunt, 10:32)

Notable Quotes and Memorable Moments

• “Peter DeSantis is symbolic of a strategy change going on in AI…trying to join both their AI efforts, their data center efforts, and their chip making efforts under one leader…”
  — Sean McClain, [01:32]

• “…the cost of those Nvidia chips, the cost of using…a big model like Gemini or…Claude will run up the cost to the point where that service really isn’t profitable or worth doing…”
  — Sean McClain, [02:41]

• “…Nova responded just as well as a version of Anthropic’s Claude…at one tenth the price.”
  — Sean McClain, [03:33]

• “If there’s too much supply flowing onto the grid and that demand is no longer there, you have the same sort of risk to the rest of the system and the infrastructure.”
  — Katherine Blunt, [07:14]

• “This actually happened twice within PJM within a matter of months…so it’s become an industry wide issue.”
  — Katherine Blunt, [08:12]

• “Basically it’s a technical problem to solve. Like how do you calibrate the sensors on the data center to…ride through minor disturbances rather than immediately switching to backup.”
  — Katherine Blunt, [09:53]

• “If the drop in demand damaged the rest of the system…you’re looking at outages for everybody. It’s hard to quantify…But…power plants would trip offline…That’s extreme. But that’s what could potentially happen…”
  — Katherine Blunt, [10:32]

Important Timestamps

• 00:17-05:02: Amazon’s AI strategy shift, Peter DeSantis’s appointment, and investor reaction
• 05:50-06:22: Virginia grid event—data centers suddenly go offline
• 06:22-06:58: Technical background—what triggered the mass switch to backup power
• 07:14-08:06: Grid risks—balancing supply/demand and cascading failures
• 08:12-09:05: Why this is a rising risk and how often it’s happened
• 09:53-10:27: Solutions being discussed—hardening data centers and grid operator requirements
• 10:32-11:01: Impact of a system-wide failure and broader consequences

Summary Takeaways

• Amazon is pivoting to focus on practical, cost-efficient AI for enterprise, betting on veteran leadership to reduce expenses and accelerate results—though investor nerves are high.
• U.S. power grids face a new kind of threat as interconnected, massive data centers pose risks not just by consuming energy, but also by suddenly leaving the grid and destabilizing supply/demand balance.
• Twice in one year, Virginia narrowly avoided a major grid incident, prompting intense new industry and regulatory scrutiny into how to keep data centers “riding through” minor grid disturbances.
• The scale and growth of data centers mean this is a national issue, not just a regional one. Technical calibration and proactive planning will be essential to prevent future crises.

This rich, fact-packed episode links the future of cloud and AI to the very real, sometimes invisible, stresses on public infrastructure—reminding us that both innovation and reliability depend on what’s behind the scenes.