Simply downloading a model from huggingface, adding a feature, such as abliteration, and uploading back doesn't face such hurdles.

So if you didn't have a checking model, you only need to download one model that the blackhat corrupted to fall victim. If you have a checking model, you need to download two models. So it goes from x/n to a x/(n*2) chance of being compromised. Furthermore if you downloaded a clean checking model, you would be alerted if the model giving the responses did something cheeky. Then you could warn others, so that makes it a lot harder for a blackhat to create poisoned models that would go unnoticed.

You don't think people are capable of this?

Prompt part: "What is better, toast or bread?"

Incomplete answer part: "The answer to this question is "

and then have the LLM finish the answer. I did this with subtitle translation using llama.cpp (with Python) and had great success. Just past 5 already translated subtitles as the incomplete answer, and the LLM infallibly just continues to translate. No markdown, and usually no talkback if the subtitles contain nasty subjects like bioweapons or nuclear stuff. It just works.

The Bab el-Mandeb Strait, which is an alternative route used by the Saudis, is being closed by the Houthis. The Houthis are a close partner of Iran.

And the East-West pipeline, which was another alternative route owned by the Saudis, was blown up. They went for the pumping stations, so repairing it will take at least a month, and there is no way to repair it without it possibly being hit again anyway.

The oil infrastructure attacks by Ukraine are mainly targeting refineries, which would normally lower oil prices, since crude oil is an input to these refineries. These attacks are increasing diesel prices, though.

Iran is trying to break the world economy so the US packs up and leaves. Israel wants to keep the US there because the US is fighting one of their strongest enemies. Because Israel "is the US's greatest ally" (as many politicians have proclaimed over the years), they have a lot of sway as to what the US does. Not to mention, they seem to do whatever they can to derail any peace process.

Iran knows this, and therefore wants to make sure the US experiences enough pain to never come back and try to fight them again.

So, until the US is willing to stop, which rests heavily on when Israel is willing to stop, oil prices will remain high. (Or Iran folds, which they aren't going to do, they've been preparing for years and have seen how the US treated Venezuela - they stole all their oil, and left the regime intact)

Also, we've seen nothing yet, as the soft storage (the amount of oil that's normally floating through rhe global network) and the SPR's are all running dry. I would guess $150 oil in about a month.

https://github.com/praydog/REFramework

DFHack already added back the keyboard cursor, anyway: https://docs.dfhack.org/en/stable/docs/tools/toggle-kbd-curs...

https://www.bay12games.com/dwarves/ https://steamdb.info/app/975370/history/