Apple AI servers unused in warehouses due to low Apple Intelligence usage

_____k 85 points 83 comments March 02, 2026
9to5mac.com · View on Hacker News

Discussion Highlights (10 comments)

abeyer

they should just scalp the RAM on ebay, that's what people actually _want_ to buy, not ai

Legend2440

What are these servers actually used for? The Siri+LLM features of Apple Intelligence aren’t launched yet, and the other features like notification summaries run on-device.

siva7

Imagine launching such a shitty product that AI servers are sitting unused in 2026.

castral

Yeah but this is how Apple has always done infrastructure/services. Their internal software teams are a mess. They constantly reinvent the wheel poorly, and then they charge a premium for exclusive access. Is anyone surprised by this?

Hasz

The rest of the FAANG has invested very heavily in cloud while Apple seems to be a laggard. GCP, AWS and Azure are all publicly available products, and cloud at Netflix, Meta seems very mature for a private offering. This is not a huge disadvantage in my opinion. Let the rest of big tech fight each other to death over cloud, while controlling a very profitable differentiated offering (devices+services). Apple keeps the M series HW out of data centers, even though it presents some very attractive performance/w and per-core numbers.

teeray

Just like Siri, it’s completely useless. I don’t need Apple Intelligence to summarize my text messages. I can skim them nearly as fast.

wizardforhire

Taken another way given apple’s enormous market reach, this could be seen as perhaps the most solid metric of actual consumer interest in ai and features ignoring hype.

righthand

Apple will utilize them when needed and scoop up extra capacity after the bubble burst.

eagerpace

I've been wondering about this failed Apple Intelligence project, but the more I think of it, Apple can afford to sit and wait. In 5 years we're going to have Opus 4.6-level performance on-device, and Apple is the only company that stands to benefit from it. Nobody wants to be sending EVERY request to someone else's cloud server.

rsynnott

I mean I think a lot of it is that they're not _really_ forcing it upon people. I think I've declined it maybe twice over the last two years. Meanwhile, Google are trying to crowbar bloody gemini in _everywhere_, and I gather Microsoft is doing ditto.

Semantic search powered by Rivestack pgvector
3,471 stories · 32,344 chunks indexed