00:11
Hello, Len.
Hello.
Alright, so we did the科学系ポッドキャスト. You seem to have forgotten that we hosted one last month.
Hey, that was off audio. That wasn't...
Well, now it is on audio.
Oh no, I'm on the record now. I did, yes. Yes, I did forget.
It's okay, it's okay.
It came back.
You had a busy beginnings of the semester type business to do, and I had a lot of time in my hands, so that's fine.
But what I try to do whenever we host, I mean, this is only our second time hosting, is to try and listen to as many episodes of the submitted episodes as possible.
One, because it's genuinely fun, and two, I also discover new podcasts that I have never heard of previously.
And also, you know, podcasts is such a weird...
Such a weird sort of creative platform, because you get very little feedback from the listeners.
There's no sort of like a YouTube comment style thing happening, or like Instagram.
You have some statistics, but it's not super reliable.
A lot of these stats is kind of ambiguous. I try not to follow too much, but like really hard to...
Just the nature of this thing, it's not a real-time interaction, and it's hard to get what we're doing.
Right, yes.
Whenever I can, I try to give people feedback, or just like, hey, this is my two cents after listening to your podcast.
Or mostly it's just like, hey, I didn't know about this.
Any feedback, I think, is like a good reminder that like, oh, there are people actually listening to what I have to say.
Yeah, you might be speaking out into an unseen audience, but the idea is that somebody is listening,
and it's nice to have the recognition that somebody is listening.
Exactly, exactly. It's not very a two-way platform, this one.
So whenever I have the opportunity, I try to do so.
But so this month, we had 23 submissions, including our own, and so that's like a lot of people.
And the general consensus that I got from people who submitted this is that Notebook, or Noto, was a pretty interesting prompt, as far as the prompt goes.
Okay, okay.
Because some people took it quite literally, like, you know, their note-taking habits, which was sort of what I originally wanted to go.
03:11
You know, do you take notes when you're listening to a seminar, or like, you know, in a meeting?
Or do you, like, would you prefer to sort of give them eye contacts and that kind of engagement?
And, or like, you know, how do you take notes? Do you take digital notes? Do you take handwriting notes?
Our episode was like, you know, some academic studies trying to get to the bottom of this, right?
Like, you know, in terms of learning ability of college students, stuff.
So that was that.
And then there's a whole lot of people I realized, they, some of them took Notebook LM, which is like the newest, hottest kid on the block, I guess, and on the AI block.
And I quite enjoyed that episode by Shibucho-san, this one.
Okay.
And he just, like, explains things really well as well, very succinctly, but very enjoyably.
At one point, he employed the unit of one Harry Potter as, like, a way to refer to the word counts of what Notebook LM can handle.
Okay.
And I forgot the exact numbers, but Notebook LM can handle, like, up to many thousands of Harry Potters in it.
That sounds roughly right to what I know about Notebook LM, so.
Yeah, but, like, you know, I love this new unit that I did not know about, you know, one Harry Potter.
He did not specify, however, which volume of Harry Potter he's talking about.
So Shibucho-san, if you're listening, I would like some specificity, because volume one, the Philosopher's Stone, has very different thickness compared to, like, the order of weeks.
I'm aware of this because I can visualize their sizing from my childhood bookshelf.
Exactly.
One looked like one could seriously harm people with it.
The other just looked like a normal children's book.
Yeah, it looked like you just pestered somebody with it.
Yeah, like, so, yeah, tell us, what was the name of this podcast again?
Shibucho-san Monozukuri no Radio.
Shibucho.
Okay, so Shibucho-san.
Is it an average, and you have simply taken, you know, something along the middle of all of the books, or have you selected one book?
We need to know.
Yeah, we need to know.
The people need to know.
A very urgent request for you.
Yep.
All right.
We mentioned at some point we would note when we were being sarcastic, right?
Yeah, this is where you will hear the bell ringing at the background.
Yeah, pretend there's a bell.
06:03
Yeah, or something a bit more alarming than a bell.
Well, that's cool.
Yeah, so this idea, some of the other stuff came in.
Yeah, exactly.
All right.
And in a similar vein, Kanemaru-san from Kagaku Plant,
Kanemaru.
Kana?
Kore wa Kagaku Plant podcast kana?
Kagaku Plant Gijutsu-sha no Plant Life tte iu podcast nandesu kedo,
he is a chemical engineer, and he talked about electronic lab notebooks.
Okay, yep.
I guess it's not the same in a sense that electronic lab notebooks so far
is not being hooked up with any large language model yet.
But he did hint about the possibility of such,
and he sort of pondered about will it make our note-taking
or recording behaviors different if we had to think about
will we eventually feed this into some kind of AI models for future learning?
And it's the question that I certainly don't have a nice,
soundbite-y answer for.
I don't think he does either.
But things that I think as people who are involved in experiments
cannot ignore these days.
As a brief sidebar,
I think what I have a problem with in that space would be
the sort of refocusing onto training the AI
instead of the problem at hand.
Now, mind you, in some context,
I did not listen to their podcast,
so I have no idea.
And it's still a very probably interesting and necessary topic,
because first off, lab notebooks,
huge, really important, electronic or otherwise.
You need to keep records of that stuff.
That's intellectual property, guys.
Yeah, it's like a really, really, really, really big deal.
Even though I have written in my experiment notebooks
some frustrating remarks about why is it not working, I'm sure.
Yeah, it becomes historical documentation at some point as well.
If I become famous,
they can dig out my PhD notebooks and be like,
hey, look, even Asami was frustrated at some point.
Yeah, there you go. That's the spirit.
Leave your frustrations for future generations.
No, don't take that away as some sort of motto.
A very unprofessional remark.
Possibly some profanity.
Yeah, it's fine. It's fine.
It's perfectly fine.
But this, the idea of like, you know,
09:02
feeding it into some sort of model, right?
On maybe the most positive outlook I can think of side,
it would be the idea that you're trying to use these,
mind you, reminder for everyone,
these are language and text-based models
that are having other things integrated into them.
Yes, the learning feature is a similar algorithm
to something like image ones, but they're not the same.
Anyway, caveat.
They take in something like a notebook page of writing and other stuff,
and you attempt to suggest possible patterns between, say,
groups in a large chemical engineering company
that don't necessarily speak to each other all the time.
However, you would have to run that by people
that are really looking at what those outputs are.
Because this is not going to, at this stage,
with any of the tech that an LLM is,
it is not going to give you,
really, it's not going to give you anything useful.
No, no, no, not at all.
And you also, this remark reminds me that
in the notebook LLM episode that Shibucho-san talked about,
he did alarm the listeners about how
AIs and language models
can straight up lie to you,
you know, no problem.
Because that's not,
there's nothing inhibiting them from telling you a lie.
And there's no sort of self-correcting factors.
Right. And there is no,
there is no lie.
Because there isn't a way for it to even know.
They don't have a concept of lie.
It's easier for us to describe it that way.
But it's not a knowing, it's not even a lie,
so much as it is just text that matched a lot of patterns,
and it does not have to be factual.
It does not have to fit reality.
Yeah, it doesn't even know anything.
No, no.
But that aside,
These are great topics, right?
Yeah, constant reminder.
And listeners, if you cannot already tell,
we love talking about this.
But it's,
yeah, it's always a good reminder that, you know,
this kind of like future,
where we rely on electronic recorded experiments,
or, you know, notebooks,
may be there in future, right?
But for that to become useful for us,
it's a lot more expensive ordeal
than we think it is right now,
just because of how user-friendly chat GPT is.
And it's, yeah,
so it's a very costly and expensive.
And Kanemaru-san also said that electronic notebooks,
12:00
for now, the way the companies are licensing,
it doesn't really make sense
unless you have a massive operation,
like a global company that shares the recipe, right?
Like, I remember my PI
sort of thought about having this electronic records for ours.
And we quickly realized that we don't have
a standardized enough procedures in our lab
for it to be useful or beneficial.
So we dropped that real quick.
Yeah, fair enough.
Yeah, so there's some sort of economy of scales here at play,
and it doesn't work for everyone.
It may work for some people,
if you have, like, you know,
your colleagues all over the world,
200 people trying to do the same experiment.
Yep.
Yeah, then maybe it's useful.
And, well, so those are there.
And then there's other episodes
where they talk more about sort of their personal habits, right?
Like, so what kind of notebooks they use,
like a brand of the notebooks that they like,
or, like, pens that they like.
There was a whole bunch of those.
And there was one about,
like, a recording diet notebook.
So it's like he, Rikedan Kenkouron,
they took the concept of notebook taking,
note taking, with their podcast theme of, like,
health practices together.
And they, like, one of them is currently
ongoing experiment,
trying what happens to his weight
as he records what he eats every day.
He's, like, not even trying to manipulate,
you know, calorie limits or protein goals
or anything like that.
He's just recording.
He is simply recording what goes into his mouth
as a way to become a bit more aware
of what he's eating and how often.
Yeah, yeah, yeah.
And so, yeah, something a little bit interesting,
I think it's sort of this type of, you know,
I don't want to give airtime to, like,
a fad diet or, like, anything like that.
But, like, I think this, like, being intentional
with what you're eating,
being, like, more aware of what you're eating
and your eating habits in general,
probably helpful for a lot of people
just to, you know, know what you're eating
or how much money you're spending
on your, you know, junk food.
Like, it happened to me the other day.
Okay. All right.
So, there's that.
And then, oh, and there was one episode
that I do want to talk about.
This one is done by Riketokuraborashiobu.
15:00
And they talked about acid-free paper.
Ah.
So, like, they went from notebook
to, like, the special notebook material
that they use for their favorite notebook
to this, like, neutral pH paper.
Right?
And, like, why it needs to be acid-free.
Or why is acid-free paper good?
Right.
I remember this as a discussion, like,
maybe even, like, 10 years ago or something.
You know, like, it was, it popped up,
but it didn't have enough,
it didn't have enough popularity to, I think,
stick in, like, the public psyche,
at least in the United States where I was.
You know, like, I remember it.
I know it is, like, a term,
but it's interesting to hear it get brought up again.
Yeah.
So, they went into, like,
an actual polymer science behind it
and talked about, you know,
fiber degradation process
and what acid plays a role in it
when you're trying to, you know,
keep your record archival quality.
And my current sort of field of expertise
being conservation science.
Yeah.
Paper is some of the more ubiquitous artist material.
You know, everything from, like,
Ukiyo-e prints to, like,
just straight-up books.
Right?
Those are all, like,
paper conservators' realm of things.
And it is really difficult
to conserve these guys, really.
It's, they come in a whole bunch of mix of fillers.
It's really hard to characterize them oftentimes.
Yeah.
And needless to say,
they're just straight-up fragile.
You know, they're papers.
Yeah.
And they, especially when it's things like books,
you know, they do come in contact
with oils on your hand and stuff.
Yeah.
So, there's so many unknown contaminants
and they become visibly yellowed
or, like, weakened.
Really difficult material to conserve.
So, it was pretty cool
that they went into that kind of, like,
degradation process about it.
So, yeah.
I mean, those are just one of the fews.
Yeah.
So, it made me think of this
and I'd be curious maybe to hear,
well, I don't know if this is, you know,
exactly where their podcast would go,
but, you know, speaking of these,
I had picked up a while ago
a notebook which is described as Limex,
which I think is just a brand name,
but it describes it as a new material
made mainly from limestone.
18:02
So, you know, I think they sold it,
probably the advertising, I think,
said stone paper or something, right?
And so, I was curious, so I picked it up.
And, I mean, it feels nice.
I don't know enough about paper
to maybe say one way or the other
what the difference is.
However, I've been so uncertain
as to what to use the notebook for
that I haven't written anything in it yet.
So, I don't actually know.
So, are we going back to like a chalkboard?
Is that what's happening?
It's not, but like, I mean, look.
So, obviously the viewers can't see,
but I mean, it's just a normal notebook.
I mean, like, it's not,
they've got roughly what I would say
the weight of, you know,
a regular piece of paper is.
A normal paper, yeah.
Maybe a bit firmer feeling.
But, again, not really sure.
I also don't know if it's only the cover
or if it's like a mix of the paper bits, right?
So, I don't quite know what that's all about.
I'm seeing the label at the back
talking about the cover being made of Limex specifically.
So, maybe it's...
Maybe it's just the cover.
Maybe it's just the cover
and maybe like normal paper.
Yeah, because, okay, a quick Google search,
Limex, is it L-I-M-E-X?
Yeah, yeah, yeah, yeah.
And it says that the main selling point of that,
Limex as a paper alternative,
is that it uses a lot less water in production.
So, you know, some elements of environmental friendliness.
They claim durability and high waterproofing features.
Okay.
And high quality textures and clear printing.
Hard to say otherwise about it
without having a sample in my hand.
Yeah.
I mean, this is the thing about most paper stuff.
You actually do have to sort of hold it
and figure out if it's going to be...
Yeah.
Yeah, but that's interesting
because it is, as a chemist,
a little bit counterintuitive
to think that limestone
is somewhat more chemically inert than paper.
Yeah.
Yeah.
And I mean, then my brain was like,
I bet you not if you put acid on it.
Right.
So, you know,
you could have a real life
baking soda volcano style thing, you know?
I mean, that might be a really good way
to erase your evidence.
I guess, yeah.
If you're a spy.
Or to reveal it.
If you want to dissolve your forged passport.
That's true, yeah.
If it's just the cover, though,
it will just reveal it
when you pour acid on it.
Oh, that's true.
Yeah, shame.
There is a company called Karst,
which I think is where
I've heard the stone paper idea before.
But I don't actually know
more of the details behind that.
21:01
So I wouldn't say, you know,
whether they're worth it
or if they're just kind of upselling themselves.
They've got what looks to be a pile
of possibly limestone
as a backboard on their story page.
But anyway.
This reminds me,
and I promise I will wrap it up,
but this reminds me that
at the beginning of grad school,
I sort of looked for my, like,
to reinvent my note taking habits.
Because I was like,
okay, this is probably going to be
the last time in my life
where I will be needing extensive notes
and note taking.
And I want
like, I felt like throughout my undergrad,
I found what works for me,
but nothing that felt like this is it.
You know, this is optimal.
And I did went into,
like, I had a short foray
with this type of, like, friction pen
induce, like, you scan
a notebook page every day
to, like, keep a digitized copy.
But the writing feel
is basically the same as your paper
because you're writing on a friction,
like, you know,
the ones that you can erase pen.
And then you write it
on a special dotted paper
so that, like,
that dotted information gets scanned
and, like, you know,
have a produced high quality PDF scans.
Yes.
It worked pretty well.
I did get annoyed that I need to,
how often I need to, like,
erase the pages.
Yeah, yeah, yeah.
But now I can confidently say
that 100% of my grad school notes
of, like, the eight courses I took
are fully digitized.
That's pretty impressive.
I'm pretty sure I have none of those.
So...
It's OK because it's not like
I look back at them
any point in my time anyway.
No, but it's, you know,
it's one of those
when somebody, you know,
digs up all of your past
when you become a famous,
you know, you know,
Yeah, yeah.
I forgot that that was written
in my future.
I become famous enough
that people want to dig
into my archive of scribbles.
They'll be able to be like,
oh, wow,
look at this crazy amount of notes
they took for the, you know,
quant class and...
Yeah, and she's still only got a B,
so...
Man, look,
everybody has trouble with this,
even the best of the best.
And there's a story
right there for you.
So...
It's OK.
But yeah, no,
it just made me think about
how different...
Like, we've been writing things
on paper for millennia, I feel,
like as human race.
Oh, maybe not quite millennia.
We've been writing it.
Millennia might be strong, but...
OK, I might have exaggerated,
but like for quite a
pretty damn long time.
And we haven't evolved much from,
you know,
24:00
essentially papyrus, right?
The sort of main ingredients
is some kind of plant fiber
that's like wetted
and pressed into one sheet,
and we write on it
with some pigments.
Yeah.
And I feel like, you know,
it's really in the past decade or so
that we figured out
how to write on things
that's other than, you know,
plant fiber sheets.
And I'm kind of interested
to sort of see how
at least my relationship evolves
with writing things.
You know, will I want to stick
to writing things on pen and paper?
Or will I become
one of those people
with Apple Pen and iPads?
No.
Or, you know,
some version of those?
Wait.
It's just so funny how...
It's hard to remember
when was the last time
I wrote things
pen and paper style?
Maybe, like,
when I went to Kyaksho
to write something
in, like, the documents?
That's probably it, right?
I have to go out of my way.
I do go out of my way
to write things down
or to note
with, like, sort of pens
on occasion.
And usually,
it's as a way
of really finalizing thoughts
or, like, keeping track of them
in a bit more of, like,
a textile way.
It feels nice.
Yeah.
I have a completely
sidebar question
that I don't know if...
Do we want to start
a new episode?
I think that might...
It might be best.
It might be nothing.
But maybe we should leave it
in another space
just in case.
All right.
Yeah.
All right.
So wrap this one up.
Let's start a new one.
Thanks, everyone,
from the Notul, you know,
themed podcast.
Thank you.
That's what we're talking about here.
We just got sidetracked.
Yeah.
I'll put the link
to the Spotify playlist
so that everyone can listen
to what everyone else
has to say about it.
All right.
Bye.
Bye.
That's it for the show today.
Thanks for listening
and find us on X
at Eigo de Science.
That is E-I-G-O-D-E
S-C-I-E-N-C-E.
See you next time.
