Yup, it works great. I actually did it myself when migrating from a centos to debian host. Worked first try, no issues (except one thing that was already broken but I didn’t know because I hadn’t accessed it recently). Containers are great for this.

This. Cloud-init, or autoinstall for Ubuntu, to get the install done, then use ansible for anything more.

Bind mounts. I’ve never bothered to figure out named volumes, since I often work with the contents outside Docker. Then I just back up the whole proxmox VM. (Yes I’m aware proxmox supports containers, no I don’t plan to convert, that’s more time and effort for no meaningful gain to me.)

You can restore that backup to a new VM. I just make sure it boots and I can access the files. Turn off networking before you boot it so that it doesn’t cause conflicts.

I don’t think you’ve looked closely enough at China, then.

“I don’t know how to run a shop, but it can’t be that hard, let’s just have AI do it!”

For a first pass, yes. I wouldn’t really trust it for an unbiased, objective perspective. Each model is only as good as its training data.

When did they ever? I remember when one of my parents got fired in the 90s, they sent the stuff from the desk in a box. Including the company desk phone!

The issue is not encryption, it’s the unauthenticated API. People can interact with your server without an account.

The DJ is the artist. Each set is an album with one track. What’s broken about that?

It could run entirely on-device.

I saw one of their Jeff Goldblum ads within the past week, so they’re still airing.

I don’t know anyone that seriously uses it. The only posts I see are corporate PR and LinkedIn “Agree?” lunatics.

Jellyfin should be fine. Why do you say it breaks it?

No. Most cameras have filters to cut non-visible light.

And any EM that passes through a mask is probably going to pass through flesh too. And any EM that’s transmitted and not reflected means it can’t be imaged by a sensor.

Very thin fabric, like a thin white T-shirt, can be transparent to IR in bright sunlight. But that’s a fairly rare case.

I heard a bit on NPR over the weekend talking about copaganda. Turns out body cams are beneficial to cops, because they can take that footage and selectively edit and release it to push a certain narrative.

If you’ve ever seen a clip on social media, it often starts a few seconds before the cop hits someone, rarely showing the full sequence of events that led up to that point.

And if they can’t edit the footage to make them look good? “Oops, we didn’t retrieve that footage in time so it was overwritten.”

“They’re too big to fail, they’re the only major domestic aircraft manufacturer!” Ok so nationalize them

It’s more powerful than the one in Windows and faster than installing a third-party one. Also it does unit conversions.

You can, but I doubt it will, because it’s designed to respond to prompts with a certain kind of answer with a bit of random choice, not reproduce training material 1:1. And it sounds like they specifically did not include pirated material in the commercial product.

The order seems to say that the trained LLM and the commercial Claude product are not linked, which supports the decision. But I’m not sure how he came to that conclusion. I’m going to have to read the full order when I have time.

This might be appealed, but I doubt it’ll be taken up by SCOTUS until there are conflicting federal court rulings.

Exemplified by this being a rant in a video with sponsorships tacked on the description and comments. Probably in the video too, but obviously I’m not going to watch it.