Huh? Where’s the Hospital?

Having a brain that still functions well enough and a ready stock of real-world experiences, I am not a frequent flier to AI destinations, but a confusion that happened (to someone else) the other day prompted me to see if ChatGPT would have been any less confused.

I was sitting waiting for a bus at the みなと赤十字病院入口 (Minato Red Cross Hospital Entrance) bus stop, when I saw a clearly confused young foreign woman looking alternately at her mobile phone and the name of the bus stop at which she had just gotten off a bus.

Seeing her confusion and that she was on crutches, I asked her where she was looking to go.

She pointed to the name of the bus stop and was wondering why she couldn’t see the hospital. Where were they hiding what was surely a substantial medical facility? The answer to the mystery lies in the way bus stops are sometimes named, and in the fact that other buses make a stop at a bus stop named みなと赤十字病院, without the 入口, which means entrance. Why does is a bus stop named XYZ Entrance not at the entrance and one without “Entrance” at the actual entrance to XYZ?

The Entrance in the bus stop name refers to it being at the entrance of the road leading to the hospital on a route that does not stop at the hospital itself.

Would ChatGPT, in all it’s wisdom, have solved this problem before forcing someone on crutches to wait for yet another bus or walk what would be a considerable and perhaps annoying distance on crutches?

My prompt to ChatGPT:

There is a bus stop in Yokohama with the name 赤十字病院入口. Where is that bus stop?

Here’s what it came back with:

The bus stop 赤十字病院入口 (正確には「みなと赤十字病院入口」) is located in Yokohama’s Naka Ward, serving the Yokohama Minato Red Cross Hospital.

[Wow, it could give me the formal name of the hospital !]

Here’s the detailed information:

  • Address: 神奈川県横浜市中区新山下3丁目15 (or nearby 3-1–6)

[I guess that’s right, but didn’t check]

[Who cares, unless you are planning on dropping a hospital buster bomb on the place?]

  • It’s situated along the bus routes that connect Yokohama / Sakuragicho / Motomachi-Chukagai stations to the hospital. [Correct, except that there is no Motomachi-Chukagai bus stop on some of the buses stopping where this occurred.] The stop is right by the hospital entrance. [Wrong. It was fooled just like the young lady on crutches. It’s nowhere near being “right by the hospital entrance.”]

ChatGPT proved that, just as a human without knowledge or experience regarding how bus stops could be named, it could mistake the name as being literally descriptive of its location.

Of course, I suspected that it would be fooled and knew the reason for the confusion.

People in the neighborhood who might frequent the hospital and have more general real-world experience riding buses (something that ChatGPT does not have and will never have) realize that the hospital entrance bus stop is not at the hospital entrance. The hospital stop is at the hospital, and the hospital entrace stop is about a five-minute walk away.

Most of them probably sense that bus stop as being right at the entrance to a  Manchinro shop, at which I frequently buy shumai, which is what I had just done when I encountered the confused visitor.

The Abnormal Requirements Many Freelancers Accept

The thought has occurred to me that, with a number of years having elapsed since the large players in the translation brokering industry transformed the lives and drastically lowered the earning power of freelance translators by imposing technology requirements, there might be freelance translators who, having experienced nothing different, think it is normal to:

  • need to use a specific software product to receive translation work,
  • need to bid on jobs on a reverse-auction click-work platform in order to get work, which these days might not even be translation work, or
  • need to log onto a specific platform and translate a document on that platform, while being monitored by the entity purchasing their efforts on that hamster wheel.

None of the above is at all normal, and all are quite foreign to relationships with direct clients that are translation consumers.

The above are merely requirements imposed on freelance translators by translation-brokering agencies that do nothing more than purchase and resell translation. Because those brokers are now in the process of moving rapidly away from ordering translation from professionals, freelancers need to quickly devise a plan B, and agencies are extremely unlikely to play a part in that plan.

The road forward is not going to be easy, but staying where you are isn’t going to end well, although it certainly will end.

Some more ideas for translators are provided on the parent website.

Microsoft’s LinkedIn Reaches Deep Enshittification

After devoting sufficient time to verifying whether my problems with Microsoft’s LinkedIn are simply misunderstandings on my part, it has become clear that the image that Microsoft projects for its LinkedIn platform is a far cry from what is actually going on behind the curtain. That, combined with the ever-increasing phoniness, fakery, and desperation of people whose posts are puked at me by Microsoft’s algorithm makes a change of approach an urgent imperative.

The Microsoft agenda is demonstrated by shadow-banning of posts.

Microsoft regularly shadow-bans posts that go counter to its business model and its agenda of promoting AI and rewarding people who promote AI. A post calling this technology out or even gently criticizing AI or its proponents will get very little reach. In fact, just about any negative post on other topics as well will receive the same reception. Cheerleading posts that are mindlessly positive are rewarded. People who want to game the algorithm to get reach are welcomed to do so. I have neither the time nor the inclination to be so inauthentic.

Phony is the new real: Faking it ’til you make it has been extended to faking it even after you make it or fail to make it.

While a large portion of LinkedIn posts have long been characterized by self-congratulatory fluff, including people proclaiming their excitement or honor about things neither worthy of excitement nor capable of bestowing honor, additionally annoying elements have been recently added to the fluffosphere on LinkedIn. One is the proliferation of charlatan coaches. Originally, these coaches were mostly claiming to teach people how to “stand out” (an overworked expression we need to retire) by creating a killer profile or posting things that attracted work.

The coaching business has now been adopted by AI prompt engineering coaches who claim you can “succeed at AI” (whatever that might mean) by learning how to give AI the proper prompts. They’ll teach you. Just sign up.

Sadly, the LinkedIn coaching business is now being engaged in by some translators looking to make money from their colleagues. They claim to teach translators how to succeed in an AI-transformed business environment. Just attend one of their paid webinars or contact them for a consultation. I have seen no evidence that these translation coaches have any intention of telling their willing victims that, having adopted AI, they will still need to acquire non-AI using clients, something which is not possible for more than a tiny portion of freelancers.

This same type of disingenuous behavior is also engaged in by translators’ organizations, some of which promote paid seminars about how to adapt to or adopt AI, aimed at freelancers for whom all the adapting and adopting in the world will not save from falling into the low-paid post-editing swamp that awaits freelancers. Without clients, AI means nothing, but translators are not told that part of the story. This is shameful behavior on the part of both individual coaches and translators’ organizations.

Posts promoting AI are so numerous that banning user accounts is meaningless.

Every day I receive countless LinkedIn posts that are suggested to me by Microsoft’s algorithm and that promote AI as the greatest thing since sliced bread; AI is the final solution to all our problems, and all you need to do is learn how to give AI the correct prompts. I had started banning accounts that made these posts, but it is clear that the account-banning approach is futile. Ban ten accounts today, and Microsoft will just come up with ten (or twenty) more tomorrow that promote AI. As Microsoft pushes forward with its agenda and business model, it becomes obvious that I should pull back from LinkedIn.

Desperation and Delusions

My feed on LinkedIn is also graced by countless posts from colleague translators delusionally denying that they can be replaced by AI, when in fact AI is already replacing large numbers of freelancers for the translation process. They also claim—correctly, of course—that AI cannot beat a good human translator, but are apparently oblivious to or unwilling to accept the reality that it doesn’t matter. There is a huge demand for translation that is good enough if cheap enough, and that demand only grows as the price drops and speed increases. No amount of complaining by freelancers is going to stop AI-using agencies they have depended on from accessing that market to eliminate the need to pay professional translators. who are left with only extremely low-paid post-editing.

All of this silliness and annoying AI promotion has brought me to the decision to post nothing more on my page on LinkedIn beyond links to content in venues I control. I might make an occasional comment on the pages of other translators or in a group, but I’m not going to donate to Microsoft any content on my LinkedIn page, which does not promote any business activity for me.

This decision is the natural result of Microsoft’s push to make AI the only topic and the only bandwagon people are urged to jump on. I am not going to jump.

The W-Word Seldom Heard from NHK News

Listening to the NHK coverage of several wars that are going on around the world, I’m wondering whether the people at NHK haven’t taken a hint from Fawlty Towers in reporting of war events. They certainly are reluctant to “mention the war,” and that includes numerous things going on in the world that are clearly and correctly referred to as wars every day by respected news sources around the world.

Wars in places such as Ukraine are referred to by NHK (in Japanese) as invasions, situations, conflicts, fighting, and other things, but almost never does NHK use the W-word.

To be fair, NHK does seem to allow non-NHK people who are being interviewed in somewhat uncontrolled and perhaps difficult-to-edit situations to use the W-word, but I have never heard it uttered from an NHK mouth in their news, and that W-word prohibition appears to extend to people who are being interviewed in an NHK studio, and who might be somehow connected to NHK, albeit via things other than NHK news; university professors come to mind. They are presumably asked to tow the non-war line, and tow it they do.

In a few online places where NHK reports the news in English, however, I have seen them use the W-word. It’s rather like “immigration,” the Japanese word for which is never used with regard to people coming to Japan to live, even permanently, although Japan has traditionally referred to its having an Immigration office. This could be the result of an unwarranted fear of the distaste the thought of immigration could evoke from the general populace.

Japanese is the only language that has any official standing in Japan. If something is said in English, it hasn’t been said, aguably doesn’t matter, and in any event can be denied by pointing to the officially recognized Japanese rendering or explanation. Japanese is the governing language, and English is provided only as a convenience; and the arrangement of deniable English is certainly convenient.

I think NHK needs to get real. Avoiding the use of the word war to describe what Russia is doing in Ukraine is not going to get Japan back its Northern Territories from Russia, and Japan was granted special permission from the the US to continue purchasing fuel from Russia in spite of the sanctions resulting from their war on Ukraine. NHK really needs to get real. It won’t cost anything, other than perhaps the loss of their reputation for being excessively careful even if it results in silly war-avoidance news coverage.

Oto-ire (音入れ) and O-toire (おトイレ): I’ve worked in both.

These two terms would appear identical (オトイレ) if written in Japanese phonetics, although they are distinguishable by an intonation difference and, of course, context. I have added hyphens to separate the semantic elements of each. These demonstrate why writing Japanese with phonetics syllabaries only is a non-starter, but that is a topic for a different post.

Oto-ire refers to the application of a sound track to a video, for example. My oto-ire (literally, sound insertion/entry) experience came with a request from a company to translate a narration of a video promoting their technology and products. They hired a professional to read my narration, and I sat in on the studio recording. I was there to handle problems in reading the narration (since it included some Japanese names) and to deal with any last-minute changes requested by my client, who was also present in the studio. This assignment provided me an informative look into how these things are done.

Fast forward a few years and we come to the toilet. My toilet experience was not working for a Japanese toilet manufacturer named after Dorothy’s dog, but we were surely not in Kansas anymore, because it required me to interpret while sitting in a toilet.

During a deposition interpreting assignment in Tokyo for patent litigation, there arose the sudden need to have a telephone conference with people in the US before the last day of depositions. A person from the Japanese party needed to participate.

We were to join in the conference from a hotel room in Tokyo. Amazingly, the telephone didn’t have even a speaker function, and we didn’t have the time to arrange for a conference-type telephone. Passing around the telephone handset while interpreting wasn’t going to work, so the solution we settled on was for me to sit in the toilet of the hotel room and interpret using the wall-mounted telephone handset provided there to enable hotel guests to answer phone calls while they were answering other calls, so to speak.

It worked fairly well, the only problem being the acoustics, which probably made me sound like I had my head in an empty barrel or perhaps in the toilet bowl.

Diversity of assignments and experiences is one of the things I enjoy about interpreting. But some interpreting jobs are garbage. One I had actually was garbage, and had me traipsing around two waste processing plants in Saitama with an environmental auditor.

You just never know where the next job will take you, but all of my subsequent encounters with toilets have been quite ordinary, and my interactions with garbage are limited to making short trips downstairs to deposit same.