You can imagine how well I slept after my exam at 3:15 in the morning. None at all, of course. It did get done, though — with criticism, which I answered with an arigatō gozaimasu — namely:
- I make a lot of mistakes with adjectives;
- I sometimes get the particles wrong;
- I speak very fast.
These are all manageable problems. There’s nothing wrong with my ability to speak. In any case, the teacher is expecting me back in September to continue the course.
As it happens, I managed to fall asleep at 5 a.m., and woke again at 12 — though truthfully I could have slept well still. In the end I went to the lab, which is a pleasant place, but don’t picture a glitzy room: functional, perfectly fit for purpose, and comfortable. The walls are white, the doors blue, and iron doors cover the labs.
Criticism on the way home
And again, just a few critical observations about things back home.
I won’t write at length now about what’s fundamentally wrong with higher education at home — maybe I’ll write a separate post about that. What isn’t unusual, though, is that a single researcher-teacher carries the whole package on their back, while others, for one reason or another, watch what happens with flippant cynicism, and even, as is the custom, comment that the carrying of the package needs adjusting. Of course someone has to carry the package. That’s how I ended up in the situation of having, this semester, 30 thesis students, 2 NKFIH projects, 3 other industrial projects, 1 role in postgraduate engineering education, several papers, and similar affairs. On top of which came finishing the doctorate, keeping up the Japan contact, and the Japanese exam.
It’s enough to test anyone. Maybe one day I’ll sit down and gather up just what my to-dos are.
The university building
Today I walked around the building, went up to the 11th floor. Nobody should expect a gleaming, hyper-futuristic laboratory — rather one where everything is there that’s needed for it to count as a lab at all and for basic research to be possible. And the students rush up and down, working constantly. The teachers — amusingly, the lab is always named after the teacher, e.g. Kato-lab, Shirai-lab — literally sit in an office facing the lab! This means that while they take big strides forward, the responsibility is right there too: that some output is continuously produced and that the lab’s system holds together. Never mind the worn linoleum on the floor, never mind the yellowed plastic, never mind that it feels like walking through a hospital — the point isn’t that, as long as everything is otherwise clean and tidy.
This, I think, is what we lack at home: proper lab practice. Yes, let people burn their hands. Yes, let them put on the coat, get it dirty, run around to see a process through. Yes, sit them down beside the serious, complex equipment, not just stroke it from a distance. On every floor of the building something is happening, even at 17:00 like now. Dazzling — I could very much be here, or somewhere like it.
Science today
It goes without saying that today I gave myself over to science instead. I try to make progress with a few papers. One is obviously the seminar. Maybe I’ll ask my labmate Sakuma what he thinks of it. He’s as crazy as I am, of course: here we sit in the lab, both working, not saying a word about anything. Actually it’s the same atmosphere at home — my office-mate is just like this — and there’s nothing wrong with that.
The other thing is that I’d like to make progress with the T-SMC papers. Two are planned, and I’d like to submit one as soon as possible. I’d thought I’d wait for the feedback from the SMC conference, but maybe I’ll do this sooner.
In the end, although I didn’t finish, I at least wrote a SISY paper. It’s still early to call the evening, but this is what I managed to put together today. Which is perhaps not so bad, given the jetlag and general fatigue. I finally asked Sakuma whether he’d like to look at my presentation slides tomorrow — since, as usual, we both worked and the time flew — but I didn’t push it on him. Of course, if there’s time, we’ll take a look tomorrow.
A criticism of science
I do have a small criticism of science, though. I looked around at how neural networks work today, and their evaluation rests rather too much on a single metric — accuracy, sensitivity. On top of that, with the handling of structured data, where it really gets interesting, it’s not even certain that only the end result matters, but rather what structure we’ve learned. I’ve started developing a metric for this.
I’m not at the end yet, but in the coming days I’ll try to share a bit more detail about it — though not necessarily every little thing, or anything from which someone could produce a paper.
Evening
You’d think that after all this not much could happen. Once again I was pleasantly surprised, even if in small things.
First: KFC. I didn’t fancy eating at Saizeriya again, and my favourite udon place was already closed, so I sat down to a KFC meal. I was pleasantly surprised. Navigating again with kitchen-table language, I successfully ordered a double-down meal with a chicken bite. And this wasn’t just a little slice, but a proper piece, the kind you’d eat in an izakaya — well, all right, it wasn’t all that tasty. But it’s true it was less greasy and lighter than at home or anywhere in Europe. Honestly, it wouldn’t be a problem generally if fast food back home didn’t always swim in grease. OK, I know, I can already hear it: “Csabi, why are you eating fast food?” Well, precisely because it’s 8 p.m. and barely anything is open. Here too the last service is around 20:30, and after that you’re on your own.
I finished eating fairly quickly, so I had time to look around a little. I found a youthful tea place, the Gongcha. I dithered a bit — I couldn’t see how late they were open — so I queued behind a young couple (a girl with dyed-blonde long hair in very baggy grey trousers and a white tee, and a taller, long-haired guy, both with typical Japanese features). The cashier guy was fairly kind, waited while I picked out my things; in the end I ordered an oolong tea with aloe vera, little ice, large size (650 ml). Then I waited, and a typically adorable Japanese girl — dyed reddish-light-brown hair, heart-shaped face, slightly narrower almond eyes — handed out the tea. The usual procedure: the tea lid sealed on, oolong poured in, and finally the straw. She seemed a little unsure, as if looking for the exact spot to drive the straw into the tea, but in the end she stabbed it in — I even cheered her on jokingly, telling her to go ahead and punch right through with some force. She smiled at it; it seemed to please her that I tried to look at the thing from a slightly different angle.
Then I moved on. All I bought was a new pillow, in case I can sleep even better. I’m very glad that prices here aren’t astronomical, and you can get by relatively cheaply — this new, comfortable pillow also cost about 3,000 yen. Nothing left after that but to go home, stretch a little, and do a bit more science.
My first Japanese ad experience
Anyone even a little versed in the Japanese pop-cultural sphere may know how bizarre and attention-grabbing a Japanese ad can be. The thing is, while Japan is genuinely a calm and quiet place, the ads — perhaps for that very reason — are colourful, bright, and, I’d venture, capable of inducing epilepsy. My first such bizarre ad experience came today at KFC. I’ll link the YouTube video (KFC × Genshin Impact), but I’ll also tell it from my own perspective, what happens in it:
I hope you enjoyed this intermezzo.
Photos
