The paper clip problem always seemed too absurd to me. Also known as the paper clip maximizer, this is the thought experiment by philosopher Nick Bostrom that imagines how a superintelligent AI with the goal of maximizing paper clip production could end up destroying the world by directing all available resources to making paper clips.
While it would be irresponsible to say this is happening, we are starting to run low on some resources. And it’s about to affect your life.
You may have heard about the global memory shortage caused, in part, by the rapid buildout of AI data centers. Just as they need semiconductors for data processing and water for cooling, these facilities need memory, or RAM, for short- and long-term data storage. Pretty much all consumer electronics, from desktop computers to smartphones, also require memory to run. The problem is that just three companies — Micron, SK Hynix, and Samsung Electronics — make almost all of the memory on the market. They can’t make it fast enough right now, and it’s unclear when they’ll be able to catch up with demand.
Normally the shortage of a computer component wouldn’t lead me to reference a thought experiment about the apocalypse, but here we are. Memory is a really important component, and as the AI data center boom sucks up more and more resources, not having enough of it means that virtually every gadget with a chip in it will either get more expensive or less innovative or both. You can think of it along the same lines as the dreaded combination of inflation and stagnation made famous by the 1970s and resurrected by the second Trump administration: stagflation. Things cost more, and they’re basically worse.
Prices are already going up, and manufacturers are already pointing to the memory shortage to explain them. What you can expect in the months, and possibly years, to come is a slowdown in the type of specification bumps you’re used to seeing in new models. (This year’s iPhone Pro 17, for example, has 12GB of RAM versus the 8GB in the iPhone 16 Pro.) You might even see manufacturers picking cheaper options for components like displays or batteries, in ways that may not be immediately obvious.
“They’re looking for anywhere to cut corners just during this timeframe to offset memory costs,” said Ryan Reith, a group vice president at the market intelligence firm IDC. He added that some companies just won’t build the higher-powered devices they’d planned to build in the near future. IDC, meanwhile, predicts smartphone sales will go down in 2026 due to the memory shortage.
There is also hoarding. There’s a veritable alphabet soup of different types of storage, but one that is essential to AI is known as DRAM. You can find DRAM in gadgets big and small — laptops, gaming consoles, TVs, cars — and as the big three memory makers direct more of that supply to AI data centers, less is available to gadget makers. So some companies are stockpiling the memory, which has the tricky effect of both driving up prices and lowering supply.
The other acronym to know here is HBM, which stands for high-bandwidth memory. This is a type of DRAM that’s specifically designed to work with the high-performance processors, like Nvidia’s Blackwell chips, that are filling out AI data centers. The profit margins on this type of memory is roughly double that of the kind of DRAM that goes into consumer gadgets, so naturally memory makers are devoting extra resources to making it, contributing to the backlog of consumer-grade memory.
This situation is going to take a while to resolve. In order to build more memory chips, memory makers need to build more factories, known as fabs, and that process takes years. Micron, for example, will soon start construction on a fab in upstate New York that won’t start producing memory until 2030. The company’s business chief Sumit Sadana told CNBC last week, “We’re sold out for 2026.”
None of this means that, if you go to the smartphone store in six months, you’re not going to be able to buy one — or that it will be twice as expensive. On the contrary, device manufacturers want to avoid sticker shock. What you’ll probably see, however, is that the price of the base model stays the same, but the components inside of it aren’t as good as they would have been. If you want the version with more memory, you’ll pay an even higher premium for those specs than you would have last year.
“It’s not over yet in terms of prices going up,” said Reith.
We don’t yet know how this ends. On one hand, the data center boom that’s gobbling up all the memory is very much tied to the AI industry, which may or may not be a bubble about to pop. On the other, the trend of rising prices spans all industries. While the rate of inflation has held steady, things cost more than they did a year ago, and they’re not getting any cheaper. If smartphone makers or laptop manufacturers realize they can sell a worse product at the same price as the better one, they might want to do that, regardless of any shortages.
When we talk about the affordability crisis, we’re not exactly talking about an expensiveness crisis. Affordable means reasonable. It doesn’t feel reasonable that the average consumer gets saddled with crappier products as the AI industry creates billionaires at a record pace. The world is not ending any time soon, but you’re probably starting to feel the effects of the shift in one way or another.
A version of this story was also published in the User Friendly newsletter. Sign up here so you don’t miss the next one!
