Something like Amazon Glacier is your best bet if you don't access data often and are okay with paying a (per-GB) monthly fee. Otherwise, you may want to build a NAS PC with high levels of disk redundancy (RAID5 or RAIDZ2)

A working nuclear thermal propulsion (NTP) engine was already built in the 60s under NASA's NERVA project. It is one of the highest technological readiness level solutions we have to the dilemma of high specific impulse versus high thrust present in the current spsce engine technologies. Imo we need something like this to make manned interplanetary missions viable.

Looks like Meteor Lake iGPU will be supported though, which is still cool for VMs with GPU accelerated desktops.

Fair, although you're probably fine for "hosting jellyfin for your family" levels even then. If cloudflare boots you, swap to a free tier Oracle Cloud VPS and set up an nginx proxy.

They removed that clause. Ctrl-f "video" on their ToS page gives 0 results

https://www.cloudflare.com/en-gb/website-terms/

Example: https://grapheneos.org/faq#baseband-isolation

Yes, the baseband is isolated on all of the officially supported devices. Memory access is partitioned by the IOMMU and limited to internal memory and memory shared by the driver implementations...Earlier generation devices we used to support prior to Pixels had Wi-Fi + Bluetooth implemented on a separate SoC. This was not properly contained by the stock OS and we put substantial work into addressing that problem.

Baseband modems were not isolated from kernel memory in stock Android, GrapheneOS had to do it themselves using the IOMMU. We do not know for sure due to the proprietary/closed-source nature of baseband modem drivers, but we have no reason to assume any OEM (Samsung, Xiaomi etc) implemented proper isolation of baseband modem and system memory.

Use argparse instead of input

AI models are universal approximators f such that y=f(x,w) with optimizable weights w that minimize some metric L(y). You can come up with a hand tuned approximator yourself that matches/beats an AI model. Does not change the fact that any approximator attempts to guess (i.e. "hallucinate") the output y based on the prior x.

Chances of Battlemage matching Navi 43?

Think it's not supported for the root filesystem anymore but that's not my usecase anyway. Still supported for non root filesystems.