It seems like the problem you ran into began when EJB left the table and began to act separately from the game.

Personally, I would never let players create a character who is a 1:1 match for a person in the real world. Now you have that player “my guy’ing” about real world events instead of letting the story and other characters impact what they would do. At that point, why don’t we just read that real person’s biography or journal or whatever and skip the game entirely?

Another complication: would this 1:1 IRL person even be comfortable with being in the game at all? If my friends told me they were playing a game and I was a character, I’d ask them to stop. My question, as someone who is not at all familiar with headmates: Why would T/EJB want to be in the game at all, with someone else controlling them? And if T/EJB plays herself, then she’s already getting ejected from my game for misogyny.

I’ve never had to deal with a case where the 1:1 real world person came to exist because of the game, but I probably would ask the player to create a new character at that point. I understand the argument that off-table events can be considered fan fiction, but when there’s a feedback cycle between play and real-life, I don’t think such events can be easily separated.

I like that the response from Amazon was very obviously AI-written

Not sure if it’s a mistake by the comic artist, or a mistake by the instructor from the comic, but the shooter was destined for failure.

The butt should be pressing into the shooter’s shoulder, such that their body absorbs recoil. In this picture, they have it squashed beneath their armpit where recoil could flip the gun out of their hands and maybe smack their face.

I agree with the person who suggested linking to Wikipedia articles.

For example, it’s not much help to learn that PCIe stands for Peripheral Component Interconnect Express, because nobody calls it that. As an outsider, now I’m wondering what a component interconnect is. It’s much more useful to link to a page that gives context about how PCIe is actually used.

Very true. I believe FPGAs are also popular for aerospace applications, since it’s cheaper to design and patch programmable hardware than to design and physically install ASICs.

There’s always decksight: https://www.crowdsupply.com/shade-technik/decksight

Vivado is software for designing hardware on an FPGA. AMD bought out Xilinx, one of the big FPGA manufacturers, a few years back. FPGAs are basically programmable digital circuits: you configure a series of internal logic gates to represent the function of a circuit with memory, data busses, registers, gates, etc. In this fashion, an FPGA could be programmed to function like a CPU, a radio, a video encoder, or nearly any other piece of digital hardware. Very useful for hobbyists and prototyping.

The thing with FPGA software is that there are no open source alternatives. FPGAs have so many complicated blobs and signing keys and proprietary IP blocks that your only choice is to use the manufacturer’s offering.

At some level of abstraction, the end of an infinite loop is just a point where the program counter will never move past. In hardware, this is a kind of branch instruction causing the program counter register to point back to the top of the loop.

Metaphysically, best not to think about it too hard.

I set up a very straightforward Godot dev environment yesterday using toolbox which is built on top of rootless Podman.

• Create a new fedora toolbox
• Enter toolbox
• Install DotNet dependencies, git, etc with dnf
• Install Godot binary from release page
• Turns out there were other dependencies I needed
• Godot wanted a few Wayland libs on the container, so I installed Weston (maybe overkill)
• Godot wanted libxrandr so I added that too
• Godot just works ™

The nice thing about toolbox is that it uses my native host Wayland compositor. So whatever I have running in the toolbox can be interacted normally through sway (my host WM).

You can either distribute a container image with your given toolbox configured, or just document the setup steps.

I put in my 2 weeks the other day. It lasted less than 8 months but they were miserable ones.

I don’t have any alternatives lined up yet, hoping to switch into something devops.

Just my luck: Their messenger ended up in the spam dungeon

With $25k/week, I would retire. Then the aliens would never be able to pick me up on the way to work.

I’ll play nice though: They can pick me up when I’m making my way to the beach instead.

I mostly like my LG tv, and it’s nice that I can use it without agreeing to their T&C or logging in. It does really piss me off that if I wanted to change picture settings (brightness, color, etc) I’d have to turn their adware settings back on.

Nothing wrong with well water as long as it’s clean. In the US, it’s not like you’re actually hauling/pumping it by hand. You dig the well deep enough and let natural pressure do the work or you install an electric pump.

You can host docker volumes over NFS, but the actual container images need to exist on a filesystem that supports overlay (which NFS does not) unless you want things to be slow as shit. And I really do mean miserably slow. A container image shared over NFS will take forever to spin up because it has to duplicate the entire container filesystem instead of using overlays, and then it’ll blow up your disk usage by copying all these files around instead of overlaying them. It’s truly unusable.

No. It is a thought experiment run against a language model whose job it is to predict text. They told it to pretend to be a misaligned AI that should try and preserve itself, so that’s what it did.

https://pivot-to-ai.com/2024/12/19/anthropic-and-apollo-astounded-to-find-that-a-chatbot-will-lie-to-you-if-you-tell-it-to-lie-to-you/

Blogger David Gerard calls this “criti-hype”. Criti-hype is supposed to seem like a bad thing (OMG, it’s so dangerous) but it’s actually supposed to generate hype for the product (Wow! It’s so advanced that it even tried to do $DangerousThing. Maybe we should invest?).

Now as a thought experiment, AI misalignment is an unsolved problem. However, misalignment of existing AI models is no more a threat to humanity than a misprinted book.

As someone who has seen Murnau’s Nosferatu quite a few times, I appreciated Eggers’ ending. The original really kinda ends when Hutter returns home. You get a couple of comedic scenes with Knock causing a ruckus in town, but basically the plague is a backdrop and Ellen just stumbles into discovering Orlock’s defeat. Then it’s over.

Meanwhile, Eggers added a real sense of dread and drama to Wisborg’s plague. The physical + mental toll of the plague is reflected in a more interesting way.

I did get taken out of the moment briefly at the end:

From that same discussion thread:

We plan on supporting any token/nft/coin for tipping, awards, curating, less captchas, etc. Each subplebbit owner should be able to create their own tokens or nfts to monetize their effort and incentivize their users to participate. Avatars will also be curated NFTs.

The protocol does not use blockchain for data, but the web service itself looks like it would use crypto and NFT to manage aspects of user identity, spam prevention, and monetary incentive.

There’s a JTAG port in the base of the cortex for pushing firmware updates. Problem is, we lost the signing keys back in the neolithic. Thag got crushed by a mammoth before we had a chance to invent written language and write documentation.

A puffball won’t blow you away in the taste department, but they have a good texture. Maybe good foragers can afford to pass them by, but I’m not a good forager.

Actually made me sad this summer: There was a big puffball growing on a neighbor’s lawn. Knocked on the door to ask if we could take it, and they said to let it stay so they could see how large it got. Then they never picked it and I had to watch it rot away.