RAMageddon

Apple quietly raised prices across its line-up overnight — no new products, just higher prices — because of a chronic, industry-wide surge in memory costs driven by AI demand. The video argues this is the "end of affordable computing," that the structure of AI spending is fragile and bubble-like, and that consumers should expect to hold onto their devices longer.

1. Apple's Price Increases

Apple took its store down briefly and returned with the same machines at higher prices worldwide, up roughly 15–20%.

Apple's stated reason: "We have never seen a component price increase this much this quickly. We've now reached the point where we need to begin raising prices."

Apple shares were down 6% the same day, whilst Micron — a memory maker — reported earnings the same week with margins up over 80%.

2. The Root Cause — "RAMageddon"

• In the first 3 months of this year, the price of RAM nearly doubled.
• Morgan Stanley says memory is up roughly six times over the past year.
• The industry nickname for the crisis is "RAMageddon."

Why it happened

• The fancy memory AI chips require is made in the same factories by the same three companies that make ordinary laptop memory: Samsung, SK Hynix, and Micron — together, essentially the entire world's supply.
• AI memory sells for far more than consumer memory, so factories have been re-pointed at AI production.
• When Micron makes one batch of AI memory, it gives up making about three batches of the normal kind — so every chip going into a data centre is roughly three that never reach a laptop.

3. The Winners

4. Industry-Wide Impact

PCs

• Dell, HP, and Lenovo have already pushed PC prices up 15–20%.
• Phone makers are quietly shipping new models with less memory than last year's, for the same price.

Gaming / Consoles (described as "hit hardest of all")

• Consoles are usually sold at a loss; that model only works if components get cheaper over time — which they no longer are.
• Microsoft raised Xbox prices again — the third increase within the last year. The Series X is now $800, and the high-storage model was killed entirely. Microsoft's cited reason: console storage and memory has gone up 2.5× since October and is expected to double again next year.
• Valve shipped the Steam Machine and had to design around the shortage in real time, using one stick of RAM instead of two, which made some games run slower.

Why there's no quick fix

• A memory factory takes years to build. New capacity doesn't come online until 2027, some not until 2028 or later.
• "You cannot negotiate your way to chips that don't exist."

5. Consumer Behaviour — "The Quietest Kind of Crisis"

The predicted response is a collective decision to stop upgrading:

• Keep your current MacBook instead of buying the next one.
• Hold your phone for another two to three years instead of trading in.
• The laptop that felt slow suddenly feels fine for another 18 months.

Anecdotes cited:

• One buyer pricing a high-end MacBook watched it jump $2,000 in a week and said, "Never mind, can't justify it."
• Another shifted their upgrade timeline from "now" to "let's get another year or two out of this thing."

Research firms reportedly expect:

• The phone market to have its worst year on record.
• PCs to drop double digits.

6. The AI Spending Picture

• The four biggest tech companies — Microsoft, Amazon, Google, Meta — are spending ~$700 billion on AI this year.
• Much of this is described as circular: Nvidia invests billions in OpenAI → OpenAI spends it renting computers from Microsoft → Microsoft invests that back into Nvidia chips. "It's a loop."
• Google announced it would sell ~$85 billion of its own stock to fund its AI buildout — described as the biggest stock sale in American history — because it doesn't have the earnings to cover the spend. Meta is reportedly doing the same.
• Of the hundreds of billions in capex going out, only ~10% is coming back as real revenue.

7. The Bubble / Fragility Argument

• No company can afford to slow down — it's a "game of chicken"; each believes that falling behind in AI means death.
• They are "wired together": if even one big player cuts chip orders, it could cascade — Nvidia sales drop, data-centre builders freeze, memory makers drown in supply, and the circular money runs backwards. "That is a row of dominoes."
• Analogy to the dot-com crash of the 2000s: it wiped out a fortune of companies but still gave us Amazon and Google — so we may be "overbuilding and building something world-changing at the same time."
• The presenter is not predicting a crash next month, but argues the structure is fragile, and that Google selling a piece of itself, Apple raising prices, and Microsoft raising prices are all signs that "the bill for AI is finally here."

8. Three Takeaways (as stated at the end)

1. Prices are not coming down — this is chronic, not temporary. If you genuinely need a machine (a work laptop, a dying phone), it may be wise to buy sooner rather than later, because the sale you're waiting for may not come. If you don't need it, keep what you've got and run it into the ground.
2. If you own tech stocks, consider diversifying and taking some profits after the long run-up.
3. Contrarian bullish call: cash (USD) — the one asset every company is scrambling to get more of to fund AI capex; everyone is cash-strapped, and markets look "a little frothy."