this post was submitted on 31 Oct 2026
39 points (100.0% liked)

Fuck AI

8622 readers
2067 users here now

"We did it, Patrick! We made a technological breakthrough!"

A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.

AI, in this case, refers to LLMs, GPT technology, and anything listed as "AI" meant to increase market valuations.

Rules:

founded 2 years ago
MODERATORS
 

TOKYO -- Fujifilm will raise production capacity for magnetic tape fivefold to meet artificial-intelligence-driven demand, marking the company's large

you are viewing a single comment's thread
view the rest of the comments
[–] Arola@sh.itjust.works 8 points 3 hours ago* (last edited 3 hours ago) (2 children)

https://archive.is/20261028093115/https://asia.nikkei.com/business/technology/artificial-intelligence/as-ai-revives-magnetic-tape-fujifilm-plans-biggest-buildup-in-two-decades

Long term storage for masses of cold data seems like it would be useful for when it rarely needs to be retrieved - my mind goes to user data like location and search terms for later use by law enforcement (but I am paranoid and all out of tinfoil). What innocuous real world uses would this data in this format be used eternal but infrequent recall?

[–] WhatAmLemmy@lemmy.world 16 points 3 hours ago

Backups. Tape is still the cheapest and most resilient option for long term storage. Considering how expensive hard drives are now, changing your backup strategy to use tape instead could save hundreds of millions vs buying the equivalent amount in hard drive storage. Add the AI cybersecurity threats, plus the likelihood of regional turmoil and data centers being targeted in warfare, and how valuable pre-AI data is, I imagine there's a desire to replicate existing backups for redundancy, and keep more of it in air-gapped cold storage.

[–] WhoIzDisIz@lemmy.today 7 points 3 hours ago

I imagine audit trails for legal purposes (both business operations, and LLM processes), storing the original data used for creating LLM training sets in case they need to go back and redo them differently, and the usual data backups.