a16zAmerica's Energy Problem: We Need A New Grid
At a glance
WHAT IT’S REALLY ABOUT
Aging U.S. grid meets surging demand, decentralization, and software opportunity
- The U.S. grid expanded rapidly in the 20th century but largely “froze” after the 1980s/90s as industrial demand and buildout capacity shifted abroad, leaving an aging, brittle system and a hollowed-out workforce.
- Interconnection delays, transformer shortages, and rising delivery costs (even as generation costs fall) are pushing large loads like data centers to bypass the grid via on-site generation and storage.
- Solar-plus-batteries are highlighted as the fastest, cheapest way to add capacity and manage peaks—Texas/ ERCOT is used as a case study—while the speakers argue for a “yes-and” energy mix including gas, nuclear, geothermal, and hydro.
- A major constraint is poor grid observability and control: limited distribution-level monitoring, weak signaling/communications, and load forecasting that still relies heavily on weather, creating market inefficiencies and reliability risk.
- Nuclear sentiment has improved (recognized as clean and dispatchable), but regulation/permitting and mega-project execution remain bottlenecks; SMRs/microreactors are framed as a resilience and defense-enabling path, especially for military use cases.
IDEAS WORTH REMEMBERING
5 ideasAmerica’s grid problem is as much institutional as technical.
The speakers attribute stagnation to workforce attrition and loss of “how to build” know-how, compounded by policy and permitting complexity that slows new generation and transmission.
Delivery is becoming the hidden driver of electricity bills.
Generation has gotten cheaper (gas/solar), but delivery costs have risen sharply due to constrained wires, aging equipment, and hard-to-procure components like transformers—making “build near load” increasingly attractive.
Co-locating power with load is becoming a default for fast-growing demand.
Data centers (e.g., Microsoft’s stance) can’t wait a decade for interconnection, so they are increasingly pursuing on-site generation and storage, shifting the grid from centralized to hybrid/distributed architecture.
Solar + batteries are positioned as the quickest reliability upgrade.
Texas is presented as proof that rapid solar buildout plus widespread batteries can add elasticity and flatten peaks faster than traditional plants, improving performance during heat waves.
Grid observability is a venture-scale gap (a ‘Splunk for the grid’).
Limited real-time visibility—especially on distribution lines—creates reluctance to connect DERs and complicates net metering; the panel argues for monitoring, analytics, and cyber/control tooling analogous to IT stacks.
WORDS WORTH SAVING
5 quotesThere is no safety, there is no national defense, there is no national security without a reliable electrical grid.
— David Ulevitch
For the last, like, 20 years, uh, effectively the grid has sort of ossified.
— Ryan McEntush
Getting a new project onto the grid today, y-y-you know, you sign up for interconnection, it could take a decade.
— Erin Price-Wright
Why aren't we deploying the world's cheapest form of power literally everywhere we possibly can, and then just putting batteries everywhere? Like, there just should be batteries everywhere.
— Erin Price-Wright
There is no Splunk for the electrical grid.
— David Ulevitch
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome