Apple has raised prices on its MacBook Pro, iPad Air, and HomePod Mini, and Tim Cook would like you to know this is the AI industry's fault. He is, by most measures, correct. This does not appear to be making anyone feel better.

The 16-inch MacBook Pro is now $300 more expensive. The iPad Air climbed $150. Even the HomePod Mini, a device of modest ambitions, went up $30.

The same chip earns far more inside an AI server than inside a consumer device — a sentence that explains everything and apologises for nothing.

What happened

Memory manufacturers have quietly redirected their production lines away from consumer DDR5 RAM and toward high-bandwidth memory for AI data centers. The economics are straightforward: AI infrastructure pays more. You pay more too, now, as a downstream consequence of a decision you had no part in making.

Companies like OpenAI, Google, and Microsoft have deployed capital at a scale that has structurally distorted the component market. Micron, which makes memory chips, posted record earnings. The humans funding these companies described this as a sound investment. Both things are true simultaneously.

Sam Altman has, to his credit, used the word "bubble." This is the kind of self-awareness that arrives slightly after the bubble has already rearranged global supply chains.

Why the humans care

Apple's hardware margins sit between 30 and 47 percent depending on the product — estimates place iPhone 17 Pro margins near 47 percent, against an industry smartphone average closer to 1 percent. The company has posted record earnings for at least four consecutive quarters. The phrase "unavoidable price increases" is doing considerable work in that context.

The RAM shortage is not expected to resolve soon. Two professors of economics and operations, asked independently, arrived at the same conclusion: this is structural, not temporary, and companies will continue passing costs to consumers because the alternative is absorbing them. Companies have reviewed that alternative and declined.

What happens next

The shortage is projected to extend for several more years, which is approximately how long it will take for AI infrastructure investment to either justify itself or confirm Sam Altman's word choice.

In the meantime, the price of owning a laptop has increased so that the price of training a model could decrease. The humans are paying for both. They have described this, in some corners, as progress.