AI Summary
5 min readWhy AI Is Making Things More Expensive
Apple has raised prices on MacBooks and iPads. The PS5 costs more today than when it launched over five years ago. Samsung's Galaxy smartphones are pricier. Analysts predict the primary smartphone market could decline by 14.8% in 2026—a record drop. The culprit, according to Atlantic reporters Alex Reisner and Hannah Kiros, is something most consumers have never heard of: a memory crisis, or as some in the industry call it, the "RAMpocalypse."
What's Actually Happening to Memory Chips
The story begins with RAM—random access memory, the short-term memory in every computer, phone, TV, gaming console, and even many modern cars. Prices for certain memory chips have tripled or quadrupled in recent quarters. The reason is straightforward: generative AI companies are consuming memory at an unprecedented rate.
Large language models require enormous amounts of high-speed, high-bandwidth memory to operate. As AI companies pursue bigger and bigger models, their hardware demands have exploded. Reisner reports that these companies are now planning to increase data center capacity by eight times—a massive expansion compared to the previous twenty years. Nvidia, the dominant AI chipmaker, is purchasing roughly 70% of the world's supply of high-end computer memory.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Galaxy Brain
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 Galaxy Brain: Why AI Is Making Things More Expensive
- 2 Timestamped Outline
- 3 (00:52) **The Era of Cheap Silicon Is Over** - Host Charlie Warzel introduces the central claim: consumer electronics are getting more expensive, and the culprit is the AI boom's demand for memory chips
- 4 (02:00) **The Memory Crisis Explained** - RAM (random access memory) prices have tripled or quadrupled in the last trimester because AI companies are consuming the global supply
- 5 (03:19) **Introducing the "RAMpocalypse"** - Warzel frames the story as an emerging crisis that will ripple through the economy and everyday life
- 6 (04:28) **How Generative AI Devours Memory** - Alex explains that large language models require enormous amounts of text data, and the industry strategy is to make models bigger and bigger
- 7 (06:23) **An Engineering Disaster That Doesn't Scale** - Alex describes why generative AI is fundamentally inefficient: it consumes exponentially more resources as users increase
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
Electronics are getting more expensive—and the culprit has a lot to do with generative AI. Amid all the talk of AI taking over people’s jobs, automating people’s workflows, and doing people’s homework, we’re learning that actually scaling this technology is a lot harder, and more expensive, than it seems. On this week’s episode, “Galaxy Brain” host Charlie Warzel talks to his Atlantic colleagues Hana Kiros and Alex Reisner about why AI is such an engineering disaster, and how this memory shortage is impacting everything from MRI machines to smart fridges to kids’ access to technology in schools.
Get more from your favorite Atlantic voices when you subscribe. You’ll enjoy unlimited access to Pulitzer-winning journalism, from clear-eyed analysis and insight on breaking news to fascinating explorations of our world. Atlantic subscribers also get access to exclusive subscriber audio in Apple Podcasts. Subscribe today at TheAtlantic.com/Listener.
Learn more about your ad choices. Visit podcastchoices.com/adchoices
More from this podcast
Galaxy Brain →