It works in reader mode.

In a class on intercultural exchange we learned that most people who move abroad (semi-)permanently go through a U-shaped curve. Starting high with a Honeymoon phase, followed by a crisis/culture shock after about 2-3 months. And then a recovery and acceptance/adjustment/mastery/adaptation/whateveryouwannacallit phase. Many people also have this process repeating in waves.

Interestingly enough, people who go through the U curve once and come out the other side usually also experience the same curve when eventually moving back to their “original” country.

When looking around where I should stay in Japan, I found many reports that APA is distributing WW2 Japanese warcrime denial pamphlets and booklets in their rooms, some written by the founder himself. I wouldn’t be surprised if they’re unhappy about documentaries tackling those topics and block them as a result.

There are also some other fun controversies about them if you look around the internet. For the record, I chose not to stay with them as a result of learning about the WW2 stuff, but all the rest didn’t really make me wanna stay there either.

I’m sure there’s a joke in there about Paraguay being on the receiving end of a lot of Nazis fleeing to South America after WW2.

I’ve actually successfully used this very flowchart to date the globe I have in my room as a hand-me-down from my grandpa.

My interpretation was that he was fainting because she looks so stunning. On second look, yours seems to be closer to the artist’s intent though.

As someone who’s learned Japanese a bunch: once you’re very familiar with the symbols, you don’t look at every little line to determine what character it is, just the general shape. The characters are built by combining a discrete and smaller set of “drawings” (called radicals). So the space of possible characters is limited to those combinations. On top of that, not every legal combination actually exists. You won’t suddenly run into 鬱, but with a different radical in the bottom left, unless you’re playing a trivia game of “spot the mistake” (which can even be difficult for native speakers, just in the same way it can be difficult for native English speakers to spell some words they’d have no trouble reading.)

I would wager some misplaced lines wouldn’t hurt readabiliity much in the same way we, in English aren’t usually struggling to read a sentence even if some of the letters are swapped/missing or a “the” is duplicated, etc. I’m sure you’ve seen examples of that before in English (or your own native language if it isn’t English).

Of course in some instances, even a tiny difference can change the meaning of a sentence entirely. This is also true both in English and logographic languages. Luckily our brains do a lot of subconscious work here too and figure out where special attention is and isn’t needed by using context and knowledge about the writing system.

(Small caveat: of course, especially in languages, there are always exceptions to every rule. And also the brain can be tricked, intentionally or not, in a variety of ways.)

What we were taught and what I’ve seen a lot in the German speaking world was “punkt for strich”, “dot before line” since the addition and subtraction symbols are written with lines and the mult/div with dots (⋅ and :).

The fact that parentheses/brackets are always top priority was taught separately (even before multiplication iirc) and once we got to powers/roots it was just quickly mentioned that they have higher prio than mult/div/add/sub.

I love you.

The “5 seconds after they started moving” is relevant. If we assume this takes place on Earth (i.e. on the surface of a sphere with a set pair of north/south poles), the angle between the two vectors changes depending on their current position.

If it’s not on the equator, it’s also slightly up to interpretation if “Due East” means they’ll turn to stay on the same latitude, always adjusting to stay moving east forever or if they’ll do a great circle. In the former case, the north moving one will eventually get stuck at the north-pole too instead of continuing their circle around the globe. Most likely not within 5 seconds though, unless the place they started was within 25 feet of the north-pole.

To actually do the math we’ll need to know (or somehow deduce) where “the place where everything about them began” is though.

At least yours were taught by actual people.

My girlfriend showed me recently that one of her profs made an AI clone of himself (voice and visual) and distributed prerecorded lessons that way. Who knows if he’s even writing the script for it. Probably not.

I haven’t dug my teeth into it as much as I should have yet, so I can’t tell you which parts of your requirements are and aren’t fulfilled. But have you had a look at Kenshi? There’s also a sequel on the horizon.

Even within Swiss German itself, the people in the Canton of Valais speak such a strong dialect (actually a group of dialects) that most of the rest of Swiss German people don’t understand them.

I wonder if they’re talking in absolute values (not smart in this case, obviously). All those listed countries are rather small, so the same % immigrants would just be a lower number…

Part of the problem here is that AI is mostly done by companies with billions of investments and in turn they NEEEEEDDDDD engagement, so they all made their AI as agreeable as possible just so people would like it and stay, with results like these becoming much more “normal” than it should or could be

I wonder how much of that is intentional vs a byproduct of their training pipeline. I didn’t keep up with everything (and those companies became more and more secretive as time went on), but iirc for GPT 3.5 and 4 they used human judges to judge responses. Then they trained a judge model that learns to sort a list of possible answers to a question the same way the human judges would.

If that model learned that agreeing answers were on average more highly rated by the human judges, then that would be reflected in its orderings. This then makes the LLM more and more likely to go along with whatever the user throws at it as this training/fine-tuning goes on. Instead of the judges liking agreeing answers more on average, it could even be a training set balance issue, where there simply were more agreeing than disagreeing possible answers. A dataset imbalanced that way has a good chance of introducing a bias towards agreeing answers into the judge model. The judge model would then pass that bias onto the GPT model it is used for to train.

Pure speculation time: since ChatGPT often produces two answers and asks the user which one the user prefers, I can only assume that the user in that case is taking the mantle of those human judges. It’s unsurprising that the average GenAI user prefers to be agreed with. So that’s also a very plausible source for that bias.

Coincidentally enough, “noon” is etymologically related to “nine”, so 10 is indeed afternoon in a very literal sense.

(Just ignore the fact that it’s originally meant to be “the ninth hour after sunrise”, so ca. 3pm)

I was taught to spell out until twelve, after that the *teen ones start, which are a lot more letters and space on the paper.

Of course it all depends on whatever the relevant style guide says. Consistency is the most important bit.

In the OOP’s example, that is solved by magecraft (which is distinct from “true” magic btw) losing potency by becoming general knowledge, thus forcing mages into working their magecraft in secrecy.

Both the mages and the Church work hard at keeping it a secret, albeit with different motives and methods.

*With AI review :)

I kept up with the drama until about a week ago so what I’m saying here is the status from back then. Someone please add any new context if I’m missing any new developments:

From what it appeared, view counts dropped but ad revenue stayed the same. Even before this whole thing, YouTube pays out for ads watched (and clicked). Pay out was not dependent on raw view count for a long time, if ever.

This suspicious behavior of view count dropping but ad revenue staying the same is actually what tipped people off that the issue was adblock related. The fact that channels with a larger focus on a younger audience seeing less of a drop also helped.

Now those view counts dropping could still have an indirect, negative effect on ad revenue, if it, e.g. automatically leads to YouTube recommending their videos less prominently.