May 5, 2026 • 17 mins
Want to hear something weird about the future of data storage?
👀 WHO ARE WE?
Weird Tech is a Weird Tech Media production and a proud member of the iHeart Network.
🎙️ Hosts: Rae Johnston and Tegan Jones
💻 Edited by Sam Blacker from the Podcast Butler
🎵 Music: More Than A Game by ColorFilm Music
🔔 FOLLOW US:
See omnystudio.com/listener for privacy information.
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:12):
Hey Tagan, Hey Ray, do you want to hear something
weird about how we store our knowledge for future civilizations?
Speaker 2 (00:18):
I assumed it was just in my memory palace, much
like animal elect of it.
Speaker 3 (00:22):
You know, if you've got another theory, sure.
Speaker 1 (00:24):
We are not relying on your memory palace for future generations.
Speaker 4 (00:27):
I'm sorry to break it to your teaga.
Speaker 3 (00:29):
Look, that's probably fair.
Speaker 1 (00:30):
Microsoft has built a system that laser's data into glass, okay,
in a way that keeps it readable for up to
ten thousand years at least.
Speaker 3 (00:42):
How do they know? Have they tested it for ten
thousand years?
Speaker 1 (00:44):
At least they have tested it under conditions that would
simulate lasting ten thousand years.
Speaker 3 (00:50):
So I don't thought you had an answer for that immediately.
Speaker 1 (00:52):
For me, I'm just like, oh, here we go, got it? No,
those straight up nos. So this is a project called Silica.
It's out of Microsoft Research, which does some really cool
fun stuff with all the surplus funds that Microsoft has,
which is a lot. It was just published in Nature,
so you know it's verifiable, peer reviewed. This isn't like
(01:13):
marketing spin. I didn't get this from a PR person.
I read that who got this from reading scientific journals?
Speaker 2 (01:17):
So they should be putting this for a PR person.
This is the story I want to know about it exactly.
Speaker 1 (01:23):
So to go back a step to the problem that
it's solving, because this problem is actually it's actually kind
of alarming when you think about it for too long,
which I did and got quite upset the way that
we currently store data, so hard drives, magnetic tapes, USB sticks,
especially all the.
Speaker 4 (01:41):
Server farms holding everything in the cloud.
Speaker 3 (01:43):
Yeah.
Speaker 2 (01:44):
Well, I was going to say, like all the things
that usually a tech that becomes obsolete within ten years
because we move on to something else. Do I still
have my self powered Western digital external hard drives that
I definitely only had very legal files on there from
lands I don't know where they are.
Speaker 3 (01:59):
About all those people that lost their crypto allegedly.
Speaker 4 (02:02):
Yeah, but it all also degrades.
Speaker 1 (02:04):
Yeah, And it's like not one of those in thousands
of years it's going to degrade things. It's like relatively soon,
like years to decades before all that physical media starts
to genuinely fail.
Speaker 4 (02:18):
And we all know this.
Speaker 1 (02:19):
This isn't a surprise that's going to creep up on
us at any time.
Speaker 2 (02:23):
Also, the other thing I guess that I think about
as I keep jumping into your story. Is that from
a historic point of view as well? Because you know,
I love this. I just have a fear that we
just end up with another Library of Alexandria situation and
the Romans are going.
Speaker 3 (02:38):
To come in and burn all of our knowledge. Maybe
not the Romans. I just had to make the history joke,
but you know, no, I do know.
Speaker 1 (02:45):
And to take a more serious note, you know, we
are currently seeing global situations where we are seeing libraries, universities, schools, hospitals,
places where records are kept. Yea, bombs destroyed, turned to rubble.
Speaker 3 (03:00):
There's book burnings that are happening.
Speaker 2 (03:02):
Not to mention that historic monuments being bombed as well,
Like it's awful exactly.
Speaker 1 (03:06):
And you know even where you look at institutions where
that sort of thing isn't happening just yet. Magnetic tape
is one of the most common ways that institutions are
archiving important data. It has a shelf life of maybe
thirty years. Under the conditions hard drives, you're looking at
ten years, which basically means every hospital, every government, every archive,
(03:31):
every library. It's like it's running on a treadmill or
something like they're constantly migrating data from dying storage to
news storage forever to keep it disappearing.
Speaker 2 (03:42):
And I can tell you that in so many cases
I can say it colloquially but with conviction, that that
wouldn't even be happening a lot of the time because
there's no time to do it. So when I was
doing my masters in museum studies, one of the jobs
that I had as one of my internships was literally
just digitizing dart, like old photos, different things. And there's
not enough time, even for getting unpaid in terns.
Speaker 1 (04:05):
To do it right, not enough time, and it's expensive
if you are getting someone to do it, who is
being paid, who is.
Speaker 4 (04:11):
On the clock.
Speaker 1 (04:13):
So we've already lost enormous amounts of human knowledge this way,
where we're continuing to lose it. So the Silica team,
this is the problem that they wanted to solve, and
they asked, what's the most durable material that we have
easy access to that we could encode information into. And
the answer to that actually surprised me at first because
(04:33):
it is glass.
Speaker 3 (04:35):
Yeah, that is surprising. I could go break some right now.
Speaker 1 (04:37):
Okay, So I think we all tend to think of
glass as fragile, but at a structural level, glass is
extraordinarily stable because it doesn't rust, it doesn't rot, it
doesn't care about moisture or you know, temperature swings that
we experience on Earth, or electromagnetic interference. We have exac
(05:00):
of glass objects that are thousands of years old and
still completely intact, which is incredible.
Speaker 3 (05:06):
Here's a question.
Speaker 2 (05:07):
I guess it got me to thinking about are we
talking about you know, one of the big things in
the last five years that we've seen phones really being
sold in is like gorilla glass because there even more durable.
Speaker 3 (05:16):
I ASSUREM, it's more durable than that.
Speaker 4 (05:19):
It's just normal glass.
Speaker 1 (05:21):
This isn't a particularly type of durable glass.
Speaker 3 (05:24):
How do you protected them?
Speaker 1 (05:25):
Well, it's quite thick, that's the thing about it as well.
So you know, we've got tempered glass, We've got you know,
big thick chunks of glass that you can't break it
unless you're really trying to. But the trick that they
had to solve was not making the glass more durable,
but it was figuring out how to write data onto
it and then read it back reliably. That was the
(05:47):
part that hadn't really been cracked before that they were
trying to narrow down on it, and they've used some
technology I'm excited about taking because I get to learn
how to pronounce a word good.
Speaker 4 (05:57):
I love it.
Speaker 3 (05:58):
Please do.
Speaker 1 (06:00):
So they use what's called a femto second laser. So
femto second means that it pulses in quadrilliance of a second,
so it's incredibly fast, it's incredibly precise, and it fires
into the glass to create tiny little structural changes called voxels.
Foxels are three dimensional pixels. They exist in the x, y,
(06:22):
and z dimensions. Awesome inside the glass, so not just
on the flat surface. And each vouxel can store more
than one bit of data, which is what makes the
density important. Yes, okay, so what you end up with
is four point eighty four terabytes of data in a
(06:44):
piece of glass that is twelve centimeters square and only
two millimeters deep. You can get that amount in there.
Speaker 3 (06:51):
That's really cool.
Speaker 4 (06:53):
Yeah.
Speaker 1 (06:53):
So for an equivalent scale, that's the researchers say that
that's about two million printed books or five thousand ultra
high definition four K films.
Speaker 2 (07:05):
I love this, and also because two millimeters doesn't sound
like much, but from a thickness perspective on glass.
Speaker 3 (07:11):
That's great. Oh, I have so many questions.
Speaker 2 (07:13):
First, I'd love to, like as a writer in a historian,
just be able to be like a huge banker in
the future, be going like, well, my worker exists in
the Y dimension.
Speaker 3 (07:21):
Thanks you. I think that would be really fun.
Speaker 2 (07:24):
I also then wonder, like, that's a lot of data
in a small surface area, or you know, in a
small bit of glass or whatever.
Speaker 3 (07:31):
How can you read this? What machine can read this?
Speaker 1 (07:34):
I have the answer for you, so to read the
data back there using machine learning to interpret the vauxhalls.
So the system that writes the data and the system
that read it are both working it together, which is
part of why this works. When previous attempts didn't work.
They couldn't quite figure that bit out. The only hitch
in this plane that I'm seeing is that this relies
(07:54):
on a future where you have access.
Speaker 4 (07:57):
To the program that wrote the data.
Speaker 1 (07:58):
Otherwise you're just trying to work out how to encode
this glass forever.
Speaker 2 (08:03):
We forgot it to save that one, we didn't write
it down.
Speaker 1 (08:13):
But to go back to your earlier question about how
they know that it will last ten thousand years. So
the shelf life estimate of this is up to ten
thousand years if stored at two hundred and ninety degrees celsius,
which is very hot. Hot of the most places on
Earth would ever get.
Speaker 3 (08:28):
So not putting that in the seed bank is what
I'm hearing.
Speaker 1 (08:31):
Well, we can because even at room temperature, we don't
have figures for that, but they believe that it would
last considerably longer because there's no degradation at room temperature.
We've only had glass storage experiments running for a few years,
not ten thousand years, so they've had to extrapolate from
all the accelerated aging tests, cooking the glass.
Speaker 4 (08:53):
Seeing when it starts to fail, and it hasn't failed yet.
Speaker 3 (08:56):
That's so cool.
Speaker 2 (08:56):
I mean, it sounds like it belongs in Jeffy's US
as ten thousand year clock mountain.
Speaker 3 (09:03):
We don't actually want that.
Speaker 2 (09:04):
I find it so fascinating because science and all of
that is not my thing. I mostly learned stuff from you,
to be frank, So for me, the idea that having
something in an extreme temperature is better for preservation is
just fascinating to me.
Speaker 3 (09:17):
That's so cool. And it's like, but we put things
in cold storage.
Speaker 1 (09:21):
Yeah, well, it's even just learning what the limits are.
It's understanding that, you know, even if we had the
heat death of the planet and it was sitting at
two hundred and ninety degrees for an extended period of time,
we might have some heat resistant alien life form come
down at some point that it does have programs that
can read this data, and they can read the entirety
(09:42):
of human existence that's etched into glass, which I think
is really cool.
Speaker 2 (09:46):
It's just a bunch of cockroaches and like Mick Jagger
and like all that are laft after a nuclear war
or something.
Speaker 4 (09:54):
Who are these people? What's going on here?
Speaker 3 (09:56):
And a bunch of history.
Speaker 2 (09:57):
But I like to think that the first one that
they would load up, for want of a better word,
is just something very stupid that they've ended up just
loading up the entire thirty or forty year backlog of
neighbors or something.
Speaker 4 (10:09):
Aha, they really need to do that.
Speaker 1 (10:11):
It's just the entire Marvel cinematic universe, and the aliens
take it as a documentary Jesus. I do need to
mention some limits on this though, So the lifetime estimates,
they don't account for mechanical stress or chemical erosion. Sure,
so if it does get dropped or it's exposed to
something caustic that does change things. The researchers are very
(10:35):
upfront about that. Also, the writing speed sixty five point
nine megabits a second is pretty fast for this kind
of system, but it's not going to change the way
that we handle active, constantly changing data anytime soon. This
is archival storage. It's not like your laptop hard drive.
We're not going to be riding onto these silicon chips
(10:57):
inside our own devices. And I think the use cases
that need things to survive legal records, medical histories, government documents,
cultural heritage archives, the kind of data that we're losing.
Speaker 4 (11:11):
It's really permanent and it's catastrophic.
Speaker 1 (11:14):
And there was a whole bunch of researchers that wrote
some associated commentary in Nature, which is I love when
they do this. Yeah, there's a whole bunch of other
researchers that have peer reviewed it, and they've jumped in
and said this is really important because of this, even
though they weren't involved in the project in any way,
they've made a comparison that is incredibly cool.
Speaker 4 (11:33):
And you might know about I was going to say.
Speaker 2 (11:36):
Like a digital version of the ark of the covenant
being close.
Speaker 1 (11:41):
They put silicon in a list with oracle bones and
medieval parchment and the modern hard drive as potential milestones
in the history of how humans store knowledge.
Speaker 4 (11:54):
So do you know about the oracle bones?
Speaker 3 (11:56):
No, I don't, please tell me about that.
Speaker 1 (11:57):
Oh, so the oracle bones their pieces of bone and
shell that ancient Chinese civilizations carved records into at around
twelve hundred BCE, and we can still read them, we
still know what they say. Also, parchment manuscripts from medieval Europe.
They've survived wars and floods and fires, eight hundred years
of general chaos happening in Europe. And the thing that
(12:20):
these formats have in common is that they're physically stable
enough that knowledge survived civilizational disruption. So glass with laser
in vauxels is making a play to be the next
version of that.
Speaker 4 (12:32):
I think that's really cool.
Speaker 2 (12:33):
I hope that I would imagine that we've learned from
history as well, that let's just assume the worst on this,
and let's say that this is a way to record
our history and archives. But from history, we also want
to make sure for these aliens or.
Speaker 3 (12:45):
Whoever, that we leave maybe a little bit of a.
Speaker 2 (12:48):
Criminals on how to read this or I'm thinking about
the Rosetta Stone, and before that was discovered, which was
in Egypt, people didn't necessarily know how to read ancient
hieroglyphics only Yeah, colors of the Rosetta Stone where it
had I believe Greek on there and a few other
languages that we were able to even decipher it.
Speaker 3 (13:07):
So we need to make sure that we can.
Speaker 2 (13:08):
Also read this in the future, because as civilization we
do change and language changes.
Speaker 3 (13:12):
And all sorts of things.
Speaker 1 (13:14):
Exactly, maybe some regular glass lasering on the door that
opens the vaults that contains all of these glass coasters.
Speaker 4 (13:22):
It's like this is season one to five of Friends.
Speaker 3 (13:26):
Right exactly.
Speaker 2 (13:27):
I do still I love and I don't love that
it can still be destroyed and instead of the Romans
coming in with fire, they're just come in with some
acid and chuck it over.
Speaker 3 (13:37):
You're like too mad.
Speaker 1 (13:39):
Yeah, I think that there are some faults with this,
but I think that they've genuinely thought of some really
important things as well.
Speaker 3 (13:46):
Yeah.
Speaker 1 (13:46):
One of the things that stood out to me as
well that it's it's just using regular.
Speaker 4 (13:51):
Off the shelf lazers.
Speaker 1 (13:52):
So it's not dependent on proprietary hardware that only Microsoft
can build.
Speaker 4 (13:57):
It's designed to be scalable and used by any one.
Speaker 3 (14:00):
I like that a lot.
Speaker 1 (14:01):
And they've also pointed out improvements that they need to
make as well, so you know, density improvements, the reading
and writing speed, getting the costs down to a point
where institutions can look at this as something that they
can use on scale. But the core tech works, and
you know, the paper's peer reviewed and Microsoft's actually be
working on this for years. I was really surprised. I
hadn't heard about it yet. Projects Silica has been running
(14:23):
since twenty sixteen. Wow, So maybe I have written about
this back in Gizmoto days and I've just forgotten about it.
Speaker 4 (14:29):
It was in the sponge that got ringed out.
Speaker 2 (14:31):
Oh no.
Speaker 3 (14:32):
But you know why we can't go and remind ourself.
Speaker 2 (14:34):
Ray because our work was deleted and not saved.
Speaker 1 (14:38):
Not saved, and we rely on the wayback machine, which
is currently being blocked by news sites all around the world,
so that AI can't use the wayback machine to scrape
articles to train their data.
Speaker 2 (14:52):
Correct And it's just the thing of like, I respect
that one hundred percent as a journalist and a writer,
but also I want to access to myself. But it's
also flawed anyway, because when you're working somewhere four years
on end and you're pumping out so many articles, it's
not even just oh, a casual one a day, which
I know to normal people, and I say that with
the highest rolem of respect, would seem also like a
(15:12):
lot we wish one a day would be lovely.
Speaker 3 (15:16):
You forget so much of what you do.
Speaker 2 (15:18):
I've forgotten it more of my like literal thousands, probably
over ten thousand articles that I wrote for that site.
Speaker 3 (15:24):
I could name a couple off my hand.
Speaker 1 (15:27):
Our own personal library of Alexandra is our brains. And
the thing that destroyed it was working in digital media.
Speaker 3 (15:34):
Yeah, yeah, like my it is not up to the task.
It just isn't.
Speaker 2 (15:39):
But I was also going to say, what I really
love about this project and someone as big as Microsoft
getting involved, is that with the last three or so
years in particular, we've seen this ramp up in just
discussion and pivot towards innovation, innovation, innovation. Obviously AI has
been part of it. It's all about the future and
(15:59):
how we can make things faster and basically just scrape
history and work just to for all of these reasons
that we've discussed before. Whether you like AI or not,
I like that there is a big tech company that
is involved also in preservation.
Speaker 4 (16:14):
Absolutely.
Speaker 1 (16:15):
I think it's something that's really important. And we are
missing enormous chunks of human history because the medium that
it was stilled on didn't survive. So yeah, the fact
that we're still doing that now with digital data, it's
just faster.
Speaker 4 (16:29):
It's more invisible.
Speaker 1 (16:30):
We've got so much more information than we ever had
access to. We've forgotten more than we've learned, is you know.
So the idea that you can get a piece of
glass the size of a coaster that could hold two
million books and outlast every civilization currently on Earth is
no small thing. So hats off to the researchers at
Projects Silica.
Speaker 2 (16:55):
And that's it for this episode of Weird Tech. If
you have any weird tech that you would like us
to come, or if some of your tech is being.
Speaker 3 (17:02):
Weird, we want to hear about it.
Speaker 2 (17:04):
You can hit us up on all social media platforms
at Weird Tech Media. We're a Weird Tech media production
and a proud member of the iHeartRadio Network. This episode
was edited by the podcast Butler. Please remember to subscribe
on your favorite podcast platforms so you don't miss the
next episode, and until next time, stay weird.
Hey Tagan, Hey Ray, do you want to hear something
weird about how we store our knowledge for future civilizations?
Speaker 2 (00:18):
I assumed it was just in my memory palace, much
like animal elect of it.
Speaker 3 (00:22):
You know, if you've got another theory, sure.
Speaker 1 (00:24):
We are not relying on your memory palace for future generations.
Speaker 4 (00:27):
I'm sorry to break it to your teaga.
Speaker 3 (00:29):
Look, that's probably fair.
Speaker 1 (00:30):
Microsoft has built a system that laser's data into glass, okay,
in a way that keeps it readable for up to
ten thousand years at least.
Speaker 3 (00:42):
How do they know? Have they tested it for ten
thousand years?
Speaker 1 (00:44):
At least they have tested it under conditions that would
simulate lasting ten thousand years.
Speaker 3 (00:50):
So I don't thought you had an answer for that immediately.
Speaker 1 (00:52):
For me, I'm just like, oh, here we go, got it? No,
those straight up nos. So this is a project called Silica.
It's out of Microsoft Research, which does some really cool
fun stuff with all the surplus funds that Microsoft has,
which is a lot. It was just published in Nature,
so you know it's verifiable, peer reviewed. This isn't like
(01:13):
marketing spin. I didn't get this from a PR person.
I read that who got this from reading scientific journals?
Speaker 2 (01:17):
So they should be putting this for a PR person.
This is the story I want to know about it exactly.
Speaker 1 (01:23):
So to go back a step to the problem that
it's solving, because this problem is actually it's actually kind
of alarming when you think about it for too long,
which I did and got quite upset the way that
we currently store data, so hard drives, magnetic tapes, USB sticks,
especially all the.
Speaker 4 (01:41):
Server farms holding everything in the cloud.
Speaker 3 (01:43):
Yeah.
Speaker 2 (01:44):
Well, I was going to say, like all the things
that usually a tech that becomes obsolete within ten years
because we move on to something else. Do I still
have my self powered Western digital external hard drives that
I definitely only had very legal files on there from
lands I don't know where they are.
Speaker 3 (01:59):
About all those people that lost their crypto allegedly.
Speaker 4 (02:02):
Yeah, but it all also degrades.
Speaker 1 (02:04):
Yeah, And it's like not one of those in thousands
of years it's going to degrade things. It's like relatively soon,
like years to decades before all that physical media starts
to genuinely fail.
Speaker 4 (02:18):
And we all know this.
Speaker 1 (02:19):
This isn't a surprise that's going to creep up on
us at any time.
Speaker 2 (02:23):
Also, the other thing I guess that I think about
as I keep jumping into your story. Is that from
a historic point of view as well? Because you know,
I love this. I just have a fear that we
just end up with another Library of Alexandria situation and
the Romans are going.
Speaker 3 (02:38):
To come in and burn all of our knowledge. Maybe
not the Romans. I just had to make the history joke,
but you know, no, I do know.
Speaker 1 (02:45):
And to take a more serious note, you know, we
are currently seeing global situations where we are seeing libraries, universities, schools, hospitals,
places where records are kept. Yea, bombs destroyed, turned to rubble.
Speaker 3 (03:00):
There's book burnings that are happening.
Speaker 2 (03:02):
Not to mention that historic monuments being bombed as well,
Like it's awful exactly.
Speaker 1 (03:06):
And you know even where you look at institutions where
that sort of thing isn't happening just yet. Magnetic tape
is one of the most common ways that institutions are
archiving important data. It has a shelf life of maybe
thirty years. Under the conditions hard drives, you're looking at
ten years, which basically means every hospital, every government, every archive,
(03:31):
every library. It's like it's running on a treadmill or
something like they're constantly migrating data from dying storage to
news storage forever to keep it disappearing.
Speaker 2 (03:42):
And I can tell you that in so many cases
I can say it colloquially but with conviction, that that
wouldn't even be happening a lot of the time because
there's no time to do it. So when I was
doing my masters in museum studies, one of the jobs
that I had as one of my internships was literally
just digitizing dart, like old photos, different things. And there's
not enough time, even for getting unpaid in terns.
Speaker 1 (04:05):
To do it right, not enough time, and it's expensive
if you are getting someone to do it, who is
being paid, who is.
Speaker 4 (04:11):
On the clock.
Speaker 1 (04:13):
So we've already lost enormous amounts of human knowledge this way,
where we're continuing to lose it. So the Silica team,
this is the problem that they wanted to solve, and
they asked, what's the most durable material that we have
easy access to that we could encode information into. And
the answer to that actually surprised me at first because
(04:33):
it is glass.
Speaker 3 (04:35):
Yeah, that is surprising. I could go break some right now.
Speaker 1 (04:37):
Okay, So I think we all tend to think of
glass as fragile, but at a structural level, glass is
extraordinarily stable because it doesn't rust, it doesn't rot, it
doesn't care about moisture or you know, temperature swings that
we experience on Earth, or electromagnetic interference. We have exac
(05:00):
of glass objects that are thousands of years old and
still completely intact, which is incredible.
Speaker 3 (05:06):
Here's a question.
Speaker 2 (05:07):
I guess it got me to thinking about are we
talking about you know, one of the big things in
the last five years that we've seen phones really being
sold in is like gorilla glass because there even more durable.
Speaker 3 (05:16):
I ASSUREM, it's more durable than that.
Speaker 4 (05:19):
It's just normal glass.
Speaker 1 (05:21):
This isn't a particularly type of durable glass.
Speaker 3 (05:24):
How do you protected them?
Speaker 1 (05:25):
Well, it's quite thick, that's the thing about it as well.
So you know, we've got tempered glass, We've got you know,
big thick chunks of glass that you can't break it
unless you're really trying to. But the trick that they
had to solve was not making the glass more durable,
but it was figuring out how to write data onto
it and then read it back reliably. That was the
(05:47):
part that hadn't really been cracked before that they were
trying to narrow down on it, and they've used some
technology I'm excited about taking because I get to learn
how to pronounce a word good.
Speaker 4 (05:57):
I love it.
Speaker 3 (05:58):
Please do.
Speaker 1 (06:00):
So they use what's called a femto second laser. So
femto second means that it pulses in quadrilliance of a second,
so it's incredibly fast, it's incredibly precise, and it fires
into the glass to create tiny little structural changes called voxels.
Foxels are three dimensional pixels. They exist in the x, y,
(06:22):
and z dimensions. Awesome inside the glass, so not just
on the flat surface. And each vouxel can store more
than one bit of data, which is what makes the
density important. Yes, okay, so what you end up with
is four point eighty four terabytes of data in a
(06:44):
piece of glass that is twelve centimeters square and only
two millimeters deep. You can get that amount in there.
Speaker 3 (06:51):
That's really cool.
Speaker 4 (06:53):
Yeah.
Speaker 1 (06:53):
So for an equivalent scale, that's the researchers say that
that's about two million printed books or five thousand ultra
high definition four K films.
Speaker 2 (07:05):
I love this, and also because two millimeters doesn't sound
like much, but from a thickness perspective on glass.
Speaker 3 (07:11):
That's great. Oh, I have so many questions.
Speaker 2 (07:13):
First, I'd love to, like as a writer in a historian,
just be able to be like a huge banker in
the future, be going like, well, my worker exists in
the Y dimension.
Speaker 3 (07:21):
Thanks you. I think that would be really fun.
Speaker 2 (07:24):
I also then wonder, like, that's a lot of data
in a small surface area, or you know, in a
small bit of glass or whatever.
Speaker 3 (07:31):
How can you read this? What machine can read this?
Speaker 1 (07:34):
I have the answer for you, so to read the
data back there using machine learning to interpret the vauxhalls.
So the system that writes the data and the system
that read it are both working it together, which is
part of why this works. When previous attempts didn't work.
They couldn't quite figure that bit out. The only hitch
in this plane that I'm seeing is that this relies
(07:54):
on a future where you have access.
Speaker 4 (07:57):
To the program that wrote the data.
Speaker 1 (07:58):
Otherwise you're just trying to work out how to encode
this glass forever.
Speaker 2 (08:03):
We forgot it to save that one, we didn't write
it down.
Speaker 1 (08:13):
But to go back to your earlier question about how
they know that it will last ten thousand years. So
the shelf life estimate of this is up to ten
thousand years if stored at two hundred and ninety degrees celsius,
which is very hot. Hot of the most places on
Earth would ever get.
Speaker 3 (08:28):
So not putting that in the seed bank is what
I'm hearing.
Speaker 1 (08:31):
Well, we can because even at room temperature, we don't
have figures for that, but they believe that it would
last considerably longer because there's no degradation at room temperature.
We've only had glass storage experiments running for a few years,
not ten thousand years, so they've had to extrapolate from
all the accelerated aging tests, cooking the glass.
Speaker 4 (08:53):
Seeing when it starts to fail, and it hasn't failed yet.
Speaker 3 (08:56):
That's so cool.
Speaker 2 (08:56):
I mean, it sounds like it belongs in Jeffy's US
as ten thousand year clock mountain.
Speaker 3 (09:03):
We don't actually want that.
Speaker 2 (09:04):
I find it so fascinating because science and all of
that is not my thing. I mostly learned stuff from you,
to be frank, So for me, the idea that having
something in an extreme temperature is better for preservation is
just fascinating to me.
Speaker 3 (09:17):
That's so cool. And it's like, but we put things
in cold storage.
Speaker 1 (09:21):
Yeah, well, it's even just learning what the limits are.
It's understanding that, you know, even if we had the
heat death of the planet and it was sitting at
two hundred and ninety degrees for an extended period of time,
we might have some heat resistant alien life form come
down at some point that it does have programs that
can read this data, and they can read the entirety
(09:42):
of human existence that's etched into glass, which I think
is really cool.
Speaker 2 (09:46):
It's just a bunch of cockroaches and like Mick Jagger
and like all that are laft after a nuclear war
or something.
Speaker 4 (09:54):
Who are these people? What's going on here?
Speaker 3 (09:56):
And a bunch of history.
Speaker 2 (09:57):
But I like to think that the first one that
they would load up, for want of a better word,
is just something very stupid that they've ended up just
loading up the entire thirty or forty year backlog of
neighbors or something.
Speaker 4 (10:09):
Aha, they really need to do that.
Speaker 1 (10:11):
It's just the entire Marvel cinematic universe, and the aliens
take it as a documentary Jesus. I do need to
mention some limits on this though, So the lifetime estimates,
they don't account for mechanical stress or chemical erosion. Sure,
so if it does get dropped or it's exposed to
something caustic that does change things. The researchers are very
(10:35):
upfront about that. Also, the writing speed sixty five point
nine megabits a second is pretty fast for this kind
of system, but it's not going to change the way
that we handle active, constantly changing data anytime soon. This
is archival storage. It's not like your laptop hard drive.
We're not going to be riding onto these silicon chips
(10:57):
inside our own devices. And I think the use cases
that need things to survive legal records, medical histories, government documents,
cultural heritage archives, the kind of data that we're losing.
Speaker 4 (11:11):
It's really permanent and it's catastrophic.
Speaker 1 (11:14):
And there was a whole bunch of researchers that wrote
some associated commentary in Nature, which is I love when
they do this. Yeah, there's a whole bunch of other
researchers that have peer reviewed it, and they've jumped in
and said this is really important because of this, even
though they weren't involved in the project in any way,
they've made a comparison that is incredibly cool.
Speaker 4 (11:33):
And you might know about I was going to say.
Speaker 2 (11:36):
Like a digital version of the ark of the covenant
being close.
Speaker 1 (11:41):
They put silicon in a list with oracle bones and
medieval parchment and the modern hard drive as potential milestones
in the history of how humans store knowledge.
Speaker 4 (11:54):
So do you know about the oracle bones?
Speaker 3 (11:56):
No, I don't, please tell me about that.
Speaker 1 (11:57):
Oh, so the oracle bones their pieces of bone and
shell that ancient Chinese civilizations carved records into at around
twelve hundred BCE, and we can still read them, we
still know what they say. Also, parchment manuscripts from medieval Europe.
They've survived wars and floods and fires, eight hundred years
of general chaos happening in Europe. And the thing that
(12:20):
these formats have in common is that they're physically stable
enough that knowledge survived civilizational disruption. So glass with laser
in vauxels is making a play to be the next
version of that.
Speaker 4 (12:32):
I think that's really cool.
Speaker 2 (12:33):
I hope that I would imagine that we've learned from
history as well, that let's just assume the worst on this,
and let's say that this is a way to record
our history and archives. But from history, we also want
to make sure for these aliens or.
Speaker 3 (12:45):
Whoever, that we leave maybe a little bit of a.
Speaker 2 (12:48):
Criminals on how to read this or I'm thinking about
the Rosetta Stone, and before that was discovered, which was
in Egypt, people didn't necessarily know how to read ancient
hieroglyphics only Yeah, colors of the Rosetta Stone where it
had I believe Greek on there and a few other
languages that we were able to even decipher it.
Speaker 3 (13:07):
So we need to make sure that we can.
Speaker 2 (13:08):
Also read this in the future, because as civilization we
do change and language changes.
Speaker 3 (13:12):
And all sorts of things.
Speaker 1 (13:14):
Exactly, maybe some regular glass lasering on the door that
opens the vaults that contains all of these glass coasters.
Speaker 4 (13:22):
It's like this is season one to five of Friends.
Speaker 3 (13:26):
Right exactly.
Speaker 2 (13:27):
I do still I love and I don't love that
it can still be destroyed and instead of the Romans
coming in with fire, they're just come in with some
acid and chuck it over.
Speaker 3 (13:37):
You're like too mad.
Speaker 1 (13:39):
Yeah, I think that there are some faults with this,
but I think that they've genuinely thought of some really
important things as well.
Speaker 3 (13:46):
Yeah.
Speaker 1 (13:46):
One of the things that stood out to me as
well that it's it's just using regular.
Speaker 4 (13:51):
Off the shelf lazers.
Speaker 1 (13:52):
So it's not dependent on proprietary hardware that only Microsoft
can build.
Speaker 4 (13:57):
It's designed to be scalable and used by any one.
Speaker 3 (14:00):
I like that a lot.
Speaker 1 (14:01):
And they've also pointed out improvements that they need to
make as well, so you know, density improvements, the reading
and writing speed, getting the costs down to a point
where institutions can look at this as something that they
can use on scale. But the core tech works, and
you know, the paper's peer reviewed and Microsoft's actually be
working on this for years. I was really surprised. I
hadn't heard about it yet. Projects Silica has been running
(14:23):
since twenty sixteen. Wow, So maybe I have written about
this back in Gizmoto days and I've just forgotten about it.
Speaker 4 (14:29):
It was in the sponge that got ringed out.
Speaker 2 (14:31):
Oh no.
Speaker 3 (14:32):
But you know why we can't go and remind ourself.
Speaker 2 (14:34):
Ray because our work was deleted and not saved.
Speaker 1 (14:38):
Not saved, and we rely on the wayback machine, which
is currently being blocked by news sites all around the world,
so that AI can't use the wayback machine to scrape
articles to train their data.
Speaker 2 (14:52):
Correct And it's just the thing of like, I respect
that one hundred percent as a journalist and a writer,
but also I want to access to myself. But it's
also flawed anyway, because when you're working somewhere four years
on end and you're pumping out so many articles, it's
not even just oh, a casual one a day, which
I know to normal people, and I say that with
the highest rolem of respect, would seem also like a
(15:12):
lot we wish one a day would be lovely.
Speaker 3 (15:16):
You forget so much of what you do.
Speaker 2 (15:18):
I've forgotten it more of my like literal thousands, probably
over ten thousand articles that I wrote for that site.
Speaker 3 (15:24):
I could name a couple off my hand.
Speaker 1 (15:27):
Our own personal library of Alexandra is our brains. And
the thing that destroyed it was working in digital media.
Speaker 3 (15:34):
Yeah, yeah, like my it is not up to the task.
It just isn't.
Speaker 2 (15:39):
But I was also going to say, what I really
love about this project and someone as big as Microsoft
getting involved, is that with the last three or so
years in particular, we've seen this ramp up in just
discussion and pivot towards innovation, innovation, innovation. Obviously AI has
been part of it. It's all about the future and
(15:59):
how we can make things faster and basically just scrape
history and work just to for all of these reasons
that we've discussed before. Whether you like AI or not,
I like that there is a big tech company that
is involved also in preservation.
Speaker 4 (16:14):
Absolutely.
Speaker 1 (16:15):
I think it's something that's really important. And we are
missing enormous chunks of human history because the medium that
it was stilled on didn't survive. So yeah, the fact
that we're still doing that now with digital data, it's
just faster.
Speaker 4 (16:29):
It's more invisible.
Speaker 1 (16:30):
We've got so much more information than we ever had
access to. We've forgotten more than we've learned, is you know.
So the idea that you can get a piece of
glass the size of a coaster that could hold two
million books and outlast every civilization currently on Earth is
no small thing. So hats off to the researchers at
Projects Silica.
Speaker 2 (16:55):
And that's it for this episode of Weird Tech. If
you have any weird tech that you would like us
to come, or if some of your tech is being.
Speaker 3 (17:02):
Weird, we want to hear about it.
Speaker 2 (17:04):
You can hit us up on all social media platforms
at Weird Tech Media. We're a Weird Tech media production
and a proud member of the iHeartRadio Network. This episode
was edited by the podcast Butler. Please remember to subscribe
on your favorite podcast platforms so you don't miss the
next episode, and until next time, stay weird.
Popular Podcasts
Betrayal Weekly
Betrayal Weekly is back for a new season. Every Thursday, Betrayal Weekly shares first-hand accounts of broken trust, shocking deceptions, and the trail of destruction they leave behind. Hosted by Andrea Gunning, this weekly ongoing series digs into real-life stories of betrayal and the aftermath. From stories of double lives to dark discoveries, these are cautionary tales and accounts of resilience against all odds. From the producers of the critically acclaimed Betrayal series, Betrayal Weekly drops new episodes every Thursday. If you would like to share your story, you can reach out to the Betrayal Team by emailing them at betrayalpod@gmail.com and follow us on Instagram at @betrayalpod and @glasspodcasts. Please join our Substack for additional exclusive content, curated book recommendations, and community discussions. Sign up FREE by clicking this link Beyond Betrayal Substack. Join our community dedicated to truth, resilience, and healing. Your voice matters! Be a part of our Betrayal journey on Substack.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com