Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

ReadMultiplex.com Podcast.

ReadMultiplex.com Podcast. artwork

ReadMultiplex.com: How the Digital Age Is Erasing Itself—And The 1000 Year Disc That Refuses to Die.

Oct 11, 20261 hr 1 minAudio

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! There is a kind of loss that does not announce itself. No fire. No headline. A tape deck stops tracking. A dye disc develops unreadable errors on its innermost tracks. A hard drive spins down in a drawer and never spins up. A photo service changes its terms, and the export path disappears. A cloud account sits dormant, invisible, until an automated policy - not a human decision - deletes it. A university lab migrates servers, and the graduate student's dataset is not on the list. For most of my working life, I have been in the business of moving records out of that kind of custody. My career began where most preservation work actually begins: not in a vault, but wherever the last working reader still sat. Payphone rooms from the late 1970s onward, where I installed and maintained machines whose only record of a call was a paper tape, a coin box, or a line card that a later owner would throw out. Garage benches. Stacks of media that had already outlived the machine that wrote them. Later, the work moved to hard drives and digital accounts. The rooms changed. The failure did not. The task was never to admire a format. It was to move a record onto something that could still be read before the window closed. The digital age does not remember everything. It generates an enormous volume of data, retains a thin slice of it, and then fails, quietly, to keep even that slice. Two numbers have to be kept apart, because they are often collapsed into one slogan. It is almost a certainty you are and will continue to lose your data: Images, videos, test and everything else, gone. But there is something you can do.. Read more at: ReadMultiplex.com Support this work by buying us a coffee: https://ko-fi.com/brianroemmele

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show. Tap a time to play from there.

98 % of all human data generated this year is, well, it's going to evaporate before the next calendar year even begins. Just let that sink in for a second. We are currently producing more information in a single day than human civilization created from the dawn of time up until like roughly 2003. And almost all of it is basically destined to just vanish. It is a staggering reality. Yeah. Now, if you are listening to this deep dive right now, you probably think you are the exception. You likely spend hours capturing these perfect

photos. You wait for the sunset to hit that exact right angle on a family vacation. And you meticulously organize those files into all these little nested folders. And you just assume that because they are floating in the cloud or sitting on this sleek aluminum hard drive right on your desk, that they are permanent. Which is such a dangerous assumption to make. Totally. Because the reality of our modern data infrastructure is that unless you take deliberate physical action, Your most precious memories and crucial records are, well,

they're quietly sitting on death row. Yeah. We really operate under this massive collective delusion when it comes to digital permanence because data feels so weightless, you know, because it moves invisibly between all our devices. We just sort of assume it has transcended the physical wear and tear that degrades physical things like paper or film. Right. Like we solved decay because it's digital now. Exactly. But digital information is entirely dependent on this highly fragile interconnected stack of physical hardware, shifting

software protocols, and let's be honest, corporate financial agreements. When we treat a digital photograph like it's carved in stone, we are fundamentally misunderstanding the medium. It's really far more accurate to view modern digital storage as an ongoing high maintenance performance that can just be interrupted at any given moment. And that is exactly why we are dedicating this deep dive to the actual mechanics of digital decay. And I mean, more importantly, digital survival. We are going to explore the architecture

of how data disappears. And then we are going to build a highly concrete, incredibly inexpensive and permanent blueprint for taking absolute custody of your digital life. Which is so crucial right now. It really is. We are talking about true self -reliance here. Something immune to corporate bankruptcies, changing terms of service, and mechanical failure. To set the scope for this conversation, though, we have to separate the signal from the noise, right? We are not addressing what network engineers call ephemeral stream

data. Okay, so what does that mean exactly? Well, we don't need a permanent archive of backend server logs or... you know, cached tracking cookies or the endless algorithmic scroll of your social media feeds. Those are temporary by design. Right. Nobody needs to archive their TikTok for you page for their grandchildren. Exactly. Our focus today is the human record. We're talking about the scanned birth certificates, the novel you spent three years writing, the audio recording of a grandparent's voice, the camera roll. We

are really isolating the data that possesses actual irreplaceable human value. The things you actually intend to keep. The core issue is that digital loss rarely announces itself. There's no modern equivalent of the Library of Alexandria burning to the ground while everyone stands around watching in horror. Digital loss happens in complete insidious silence. It's so quiet. Right. You only discover the loss at the exact moment you try to retrieve the file, and it simply isn't

there. And this brings up a really fascinating framework from our sources for understanding why this happens. Any digital file is essentially governed by two competing countdown clocks. Yeah, the two clocks framework is essential here because it divides this illusion of digital permanence into two very distinct factors of failure. So the first is what we call the mediums clock, which represents pure physics and mechanical engineering. OK, break that down for us. Well, information has to be inscribed onto a physical

substrate, right? And that substrate degrades over time. Take magnetic tape, for instance. Tape was the absolute gold standard for enterprise backups for decades. But magnetic tape is governed by coercivity. Coercivity. Yeah, it's the tendency of the magnetic particles to just lose their charge over time or be completely messed up by ambient magnetic fields. And furthermore... Tape requires an incredibly complex, fast -moving mechanical tape deck just to be read. Right.

And if the manufacturing of those specific tape decks stopped like 15 years ago, the tape itself just becomes a plastic brick regardless of whether the magnetic charge on it is actually still intact. Precisely the issue. And we saw this heavily with microfilm and microfiche, which were heavily relied upon by libraries and institutions between roughly 1870 and 1970. The medium's clock for microfilm involves the physical decay of the film base itself, sometimes resulting in what archivists call vinegar syndrome. Vinegar syndrome.

It sounds awful. It is. It's where the cellulose acetate degrades and literally off -gases acetic acid, so it smells like vinegar. But the clock also involves the optics. You need a very specific reader, a really precise focal lens, and a high -intensity halogen lamp just to project it. So the hardware to read the medium fails before the medium itself. Exactly. And a modern hard drive is no different. We view them as these magic black boxes on our desks, but a hard drive is basically a microscopic jet engine. A jet engine. I mean, think about it. It consists of

multiple metallic platters spinning at 7 ,200 revolutions per minute, with a reed right head flying mere nanometers above the surface on a cushion of air. It requires these tiny fluid dynamic bearings that must not seize up, and a delicate actuator arm that absolutely must not crash. The medium's clock is relentless because basic thermodynamics and friction guarantee that complex mechanical systems will eventually fail. Okay, so that physical clock makes total sense. Everything physical wears out eventually. But

it's this second clock that feels much more treacherous for the average person today, which is the access points clock. Right. So the access points clock is entirely administrative and bureaucratic. In our current era, most people do not hold their data. They hold a key to a remote server. Like iCloud or Google Drive. Exactly. A cloud object depends on an unbroken chain of administrative realities. It requires the hosting corporation to remain solvent. It requires the service contract or the end user license agreement to remain totally

unchanged. It relies on API endpoints remaining active, a valid login credential that hasn't been flagged by some automated security algorithm. and a credit card that has not expired. That's the scary part. If you are hospitalized for three months, your credit card expires, and you miss the warning emails because you're incapacitated, that entire administrative pillar collapses. The file hasn't physically degraded. The magnetic charges representing your family photos are technically still sitting on a server rack in an Amazon or

Google data center somewhere in Virginia, but they are gone to you. Archivists actually refer to this as dark data. The bits exist on a physical platter somewhere. But functionally, the information has vanished from the human record because the administrative bridge to reach it has just been severed. And we have massive historical precedent showing how quickly this administrative clock can run out on a truly global scale. Oh, man. MySpace is the textbook example of this, isn't it? When we look back at the early 2000s, MySpace

was the epicenter of global Internet culture. Millions of musicians used it as their primary and often their only distribution platform. Yeah. And then in 2019, MySpace issued a rather quiet statement confirming that a server migration had resulted in the catastrophic loss of essentially all music uploaded to the platform between 2003 and 2015. It's just mind -blowing. We're talking about the complete erasure of roughly 50 million songs from 14 million artists. Just gone. But

wait, how does a basic server migration wipe out 50 million files? You would assume a tech giant has redundancies upon redundancies for that sort of thing. You would assume so. But in these complex database architectures, the data is often separated from the pointers that locate it. So during a massive migration from old storage arrays to new ones, if the database schema linking the user profiles to the specific file paths on the storage volumes becomes corrupted, those files become orphaned. Oh, wow. So the

system literally forgets where the files live. Exactly. Even if the underlying audio files were temporarily still on the old servers, if the index is destroyed and the old servers are subsequently wiped or decommissioned before the engineering team catches the error, the data is irreversibly gone. MySpace's later advice to users to basically keep your own backups was this grim admission that the platform was never an archive to begin with. It was merely a temporary hosting service. It wiped out an entire generation's cultural

output. Just vanished. And then there was the photobucket situation. In 2017, they initiated what essentially amounted to a hostage situation. Because photobucket was the backbone of image hosting for thousands of early internet forums. eBay listings and personal blogs, right? People would upload a photo there and embed the link somewhere else. Right, and overnight, photobucket just... altered their terms of service, completely prohibiting free third -party image hosting. They replaced millions of embedded images across

the entire public web with a generic massive banner demanding that users pay a $399 annual subscription to unlock their photos. It's wild. Relying on modern digital storage under these conditions feels exactly like renting an apartment where the landlord holds the absolute legal right to just arbitrarily change the locks while you were at work. They can suddenly triple the rent, or they can just demolish the entire building while you're on vacation. You don't possess the space where your memories live. You are just

visiting under their temporary grace. The landlord analogy is perfectly accurate. You are merely a tenant in the cloud ecosystem. If publicly traded corporations with massive engineering teams are accidentally or... intentionally wiping out millions of cultural artifacts, either through sheer negligence during a migration or through aggressive monetization strategies, we have to seriously examine the infrastructure we all rely on every single day. We are experiencing what data scientists call the great divergence. Okay,

the great divergence. This is the widening gap between the sheer volume of data we generate as a species and the amount we actually retain. I really want to dig into the scale of this because the numbers are almost hallucinatory. The International Data Corporation, the IDC, tracks the global creation of data. And their projection for the year 2025 is that human civilization will create 218 .4 zettabytes of data. Which is a number so large it loses meaning. To contextualize a

zettabyte, it really helps to step up the ladder. So a gigabyte is roughly a symphony in high fidelity audio. A terabyte is a large physical library of books. A petabyte is the entire digital archive of the Library of Congress. An exabyte is a thousand petabytes. We are currently generating roughly 600 exabytes of data every single day. 600 ,000 times the digital footprint of the Library of Congress every 24 hours. That is genuinely hard

to wrap my head around. But the critical metric here is retention. The IDC reported that in 2020, less than 2 % of the data created or replicated was actually saved into the following year. So 98 % of everything digital evaporates almost instantly. Now, I have to challenge this slightly. I hear that 2 % retention number, but I look around. We have more server farms being built than ever before in history. Hyperscale data centers are consuming significant percentages of national power grids. Apple, Google. Microsoft,

they offer terabytes of cloud storage to consumers for the price of a cup of coffee. We clearly do not lack the physical capacity to store data. So why is the web decaying so rapidly if we have seemingly infinite room to store it? Why is link rot a worsening crisis? Because while we have the capacity, the architecture of the web itself is fundamentally antithetical to permanence. We really have to examine the nature of a hyperlink. A link is not a document. It is not possession.

A web link is nothing more than a temporary string of text pointing to a specific directory on a specific server, sitting in a specific geographic location operated by a specific company under a very specific financial model. So it's a pointer, not the thing itself. Exactly. When you click a link, your browser makes a DNS request to resolve an IP address and basically asks that server to hand over a file. If that company goes out of business, if they rebrand and change their URL structure, or if a systems administrator

simply moves a file from one folder to another without setting up a proper redirect, that citation resolves to a 404 error. Link rot is the inevitable result of treating temporary network addresses as permanent historical citations. The page was never in the reader's possession to begin with. You essentially memorize the driving directions to a library. But the city bulldozed the library to build a parking lot. The directions are perfect, but the destination is just gone. And the statistics on this web decay from our sources are deeply

unsettling. A 2024 Pew Research study analyzed millions of web pages. They found that 38 % of web pages that existed in 2013 were completely eradicated a decade later. And the acceleration of that decay is the truly alarming part. Because the same study found that pages created in 2021 were going missing at a rate of roughly one in five within just two years. One in five? Yeah. We are losing 20 % of our immediate digital history

almost as fast as we generate it. And this isn't just affecting obscure personal blogs from 2005 or defunct local businesses. It is eroding the foundational pillars of our shared knowledge. Just look at Wikipedia. We treat Wikipedia as the modern, indestructible library of Alexandria. But Wikipedia is entirely built on citations. Without citations, it's just a collection of unsourced claims from anonymous users. And yet, 54 % of Wikipedia pages contain at least one

reference link to a webpage that no longer exists. And the epistemological crisis is even more severe in academia and medicine. A recent analysis demonstrated that the inaccessibility of web references in medical journals rises to over 41 % within five years of publication. That's terrifying. Think about the implications of that for a second. A peer -reviewed medical paper is published, making a vital claim based on external data sets or previous online research. Five years later, nearly half of the evidence supporting that paper

cannot be verified by a new researcher because the links are simply dead. The foundational bedrock of scientific reproducibility is crumbling because it relies on temporary URLs. But wait, surely the Internet Archive and the Wayback Machine are solving this, right? That's their entire mandate. I mean, if I hit a dead link in an article, my first instinct is to immediately plug it into the Wayback Machine to find the cached version. The Internet Archive is an invaluable resource, obviously, but it is not a panacea. And it is heavily constrained by the administrative clock

we talked about earlier. A comprehensive study analyzing 27 .3 million archived URLs within the Internet Archive found that only 35 .3 % were still accessible and alive by 2023. Really? Why is that number so low? If the Archive crawls a page and saves it on their servers, shouldn't it be permanent? It comes down to how the modern web is constructed. In the early 2000s, web pages were mostly static HTML text documents. A web

crawler could easily ingest and save the entire page. Today, the web is heavily reliant on complex JavaScript frameworks, dynamic database calls, and highly personalized content delivered through APIs. A web crawler often cannot bypass a login wall. or it simply cannot execute the complex JavaScript required to render the text. Oh, so it just captures a blank loading screen, essentially. In many cases, yes. Furthermore, what we call deep links, URLs that point directly to a specific

PDF, a high -resolution image, or a subdirectory dataset rather than a top -level homepage, have a terrifyingly short lifespan. The half -life of a deep link is just one year. One year. Only about 4 % of deep links survive beyond a decade. The archive cannot catch everything. And even when it does, it often only catches the visual facade of the page, not the deep data behind it. This paints a really grim picture of the public Internet. It is essentially an untended garden built on transient financial contracts.

But I know what you, the listener, are thinking right now. You were thinking, OK, sure, the public web is rotting and Wikipedia links are breaking. But my personal data is different. I pay for a premium Google One subscription. I have a beefy two terabyte external hard drive sitting right next to my laptop. I use Apple's time machine. My home ecosystem is perfectly curated and safe. And that false sense of security is actually the most dangerous element of personal data management.

To dismantle that illusion, we have to shift our temporal perspective entirely. We have to look at what data scientists call the 25 -year horizon. The 25 -year horizon. This is brilliant because it forces you out of the day -to -day mindset. 25 years is roughly one working lifetime. It's the span between the birth of a child and them graduating from college. So let's look back exactly 25 years from today, back to roughly 2001. Think about the hardware and the storage

mediums you were using in 2001. We were relying on Iomega zip disks. We were using mini DV tapes to record home videos on camcorders. We were carefully burning CDRs using massive desktop towers. The early digital cameras used compact flash cards that held like 16 megabytes of data. Right. Now ask yourself, how much of that 2001 media can you seamlessly mount and read on the ultra -thin laptop you are using today? You likely do not have a SCSI port for the zip drive. You

definitely do not have a FireWire 400 port to ingest the mini DV tape. Most modern laptops haven't had a built -in optical CD drive in over a decade. Yeah, I'd have no way to plug any of that in. Exactly. You might be able to recover some of that data by hunting down specialist legacy hardware on eBay, buying old cables, and running virtual machines with legacy operating systems. But for the average consumer, that data is functionally unreadable. One working lifetime completely obliterates consumer digital media. It is not a theoretical risk. We have already

lived through the empirical proof of it. And the reason standard backup methods fail over this 25 -year timeline is due to a mathematical concept that completely shattered my confidence in my own setup when I read the source material. It is called the compounding arithmetic of failure. Let's break this down, starting with the cloud, because people just assume that Google or Amazon's massive multi -billion dollar data centers offer immunity to failure. The hardware in a Google data center is indeed highly redundant, but your

access to it is subject to the administrative clock. Let's look at Google's standing policy. As of recent updates, a personal Google account that remains inactive for two years may be entirely deleted. Entirely? Yes. That includes everything stored in Gmail, Google Drive, and Google Photos. So if you pass away, or if you suffer a severe cognitive decline and forget your password, or you are simply offline for a protracted period like an illness, your absence triggers an automated corporate deletion policy. It's not a server

crash. It's literally a script running on a schedule designed to free up server space. Exactly. It's just a cleanup routine. And even if you remain active, what happens when you want to migrate your data out of the cloud? Say the subscription price quadruples over the next decade. Cloud providers utilize what are called egress fees. They make it incredibly cheap, often completely free to upload your data. But when you want to download terabytes of your own files to move to a competing service, they charge you exorbitant bandwidth fees. So they trap you. They financially

trap you inside their ecosystem at the exact moment that leaving becomes necessary. The cloud is a hotel. It is not a vault. Okay, so the cloud is vulnerable to administrative failure and financial capture. What about local hardware? I have an external hard drive sitting quietly in my desk drawer. It is protected from the elements. It's not being bumped around in a backpack on the subway. Surely a hard drive that is safely stored away is a permanent archive. You'd think so, but let's look at the failure rates of mechanical hard drives. Backblaze is a major cloud storage

and backup provider, and they publish highly detailed reliability statistics on their massive fleet of hard drives. Their data shows that even enterprise -grade drives running continuously in climate -controlled, vibration -mitigated server racks fail at an annualized rate of around 1 .36%. Which doesn't sound terrible for a server farm. No, but a hard drive sitting unpowered in your desk drawer is mechanically in a far worse position. This counterintuitive fact amazed me. You assume that turning a device off preserves

its lifespan. How is an idle hard drive worse off than one spinning 24 -7? Because a hard drive is fundamentally mechanical. Inside that sealed casing, the platters spin on bearings lubricated by specialized fluids. When a drive sits unpowered for years, those fluid lubricants settle and degrade. The mechanical tolerances are so microscopic that when you finally plug that drive in after five or ten years of sitting idle, the spindle motor often lacks the torque to overcome the stiction, the static friction of the settled

bearings. Oh, wow. So it just seizes up. The drive simply refuses to spin up. Or worse, the reed right heads stick to the platters and physically scratch the magnetic surface, destroying the data instantly. A hard drive is designed for constant operation. Three to five years is a responsible, reliable operational lifespan. Planning to use a hard drive as a 25 -year archival vessel is an engineering fallacy. Okay, but what about SSDs? Solid -state drives. They have completely taken over the market. They are small. They are

incredibly fast. And crucially, they have zero moving parts. There are no bearings to seize up. I use external SSDs for everything now. Unpowered consumer SSDs are arguably even more precarious for long -term cold storage than mechanical drives. We have to understand how NAND flash memory, which is the core technology of an SSD, actually stores data. An SSD does not use magnetism. It stores data by trapping microscopic pools of electrons inside billions of microscopic cells, specifically within a component called a floating

gate transistor. So it's essentially holding an electrical charge in a tiny microscopic cage. Correct. But those cages are not perfectly impermeable. Over time, due to the laws of physics and ambient thermal energy, those trapped electrons slowly leak out through the oxide insulation layer. If enough electrons leak from a cell, the controller can no longer determine if that cell represents a digital one or a zero. The data corrupts. When an SSD is plugged into a computer and powered on, the drive's controller constantly runs background

algorithms to check cell voltage and rewrite data to refresh the charge. But if it's sitting in a drawer? If it is unpowered, that refresh cycle never happens. The official standards body for microelectronics, JEDEC, publishes client requirements for SSD retention. The standard guarantees a minimum of only one year of data retention for a consumer drive powered off. and stored at 30 degrees Celsius, which is about 86 degrees Fahrenheit, essentially the temperature of a warm room or a closet in the summer. One

single year. Wait, people are buying $300 SSDs, loading their entire family's video history onto them, tossing them into a fire safe in the garage, and assuming they are good for decades. It is a qualification floor, meaning the drive might last longer, perhaps. three to five years, depending on the exact NAND architecture and storage temperature. But the physics remain unchanged. Thermal energy guarantees electron leakage. An SSD is a brilliant high -speed working drive. It is a catastrophic

long -term archival medium. So let's bring this back to the compounding arithmetic of failure. If the cloud, mechanical hard drives, and solid -state drives all carry these inherent annual risks, what does the math actually look like over a 25 -year horizon? The mathematics of probability compounding are just unforgiving. Let's assume you have a storage strategy. Maybe it's a cloud account or an external drive that faces just a 4 % chance of a fatal event in any given year. A 4 % risk means there is a 96 % chance your

data survives the year intact. That feels incredibly safe, right? Yeah, 96 % sounds great. But over 25 years, you must multiply 0 .96 by itself 25 times. When you do that, the chance of your archive surviving the full quarter century drops to approximately 36%. A 4 % annual risk translates to nearly a two -thirds chance of total catastrophic loss over a working lifetime. And 4 % is an incredibly optimistic baseline. If we factor in typical

household chaos, spilled coffee, a stolen laptop, an expired credit card, a forgotten cloud password, a 10 % annual risk is a much more realistic assessment. At a 10 % annual risk, the survival rate of your data over 25 years is just 7%. 7%. That means catastrophic failure isn't just a possibility. It is a statistical near certainty. You don't need a house fire to lose everything. You just need one bad year out of 25 to wipe out the other

24. Now, I want to introduce a counter -argument here, because I know the tech enthusiasts listening are waiting for this. What about a NAS, a network -attached storage device? Many serious photographers and tech -savvy people have a multi -bay NAS sitting in their home network closet. They run RAID arrays, redundant array of independent disks, where the data is automatically striped and mirrored across multiple hard drives. If one drive physically breaks, the data is perfectly mirrored on the others, and you just slot in a replacement drive. Isn't that the ultimate foolproof safe? This

is perhaps the most vital distinction in data management. A NAS running a RAID array provides availability, not archivability. Break down the mechanics of why that matters. RAID architecture is designed to solve one specific problem. Uninterrupted uptime in the event of a single mechanical component failure. If disk number two in a four -disk array suffers a read -write head crash, the RAID controller uses parity data or mirroring on the other disks to seamlessly keep your files accessible while you replace the broken drive. That is availability.

However, RAID does absolutely nothing to protect you against systemic or environmental failures. Systemic failures like what? If your home network is breached and a ransomware payload executes, it will instantly encrypt the primary drive and immediately mirror that encryption across every redundant drive in the array. Your data is perfectly, redundantly destroyed. If a power surge circumvents your surge protector, it can fry the entire chassis and every drive inside it simultaneously. Or, as the scenario goes, if a pipe bursts in your

ceiling while you're at work and floods the office, the NAS enclosure, the redundant drives inside it, and the router controlling it are all sitting in the exact same puddle of water. Precisely. A NAS is a phenomenal tool for a working professional who cannot afford an hour of downtime while editing a video project. But it is not an archive. An archive is a medium that survives completely independently of your current electrical, network, and hardware infrastructure. Okay, so if we summarize everything up to this point, the public web is

a fragile facade, cloud services are hostile to long -term ownership, mechanical hard drives seize up, SSDs bleed electrons, and expensive NAS setups are just redundant vulnerability. It sounds like an utterly impossible, overwhelming crisis. I can imagine you, the listener, staring at your laptop right now, paralyzed by the sheer volume of data you think you need to save and the apparent lack of any viable medium to save it on. But this is the exact moment where this

deep dive pivots from outlining the problem to delivering an incredibly liberating solution. Because the anxiety we feel is mostly based on an illusion. What if the amount of data that genuinely matters, the human record we discussed earlier, is exponentially smaller than we fear? This is the concept of the camera roll scale. The primary source of data anxiety is the illusion of massiveness. We look at our laptops and phones and see terabytes of consumed storage. We assume

our personal legacy is petabytes of data. But when you analytically strip away the digital noise, the reality is astonishingly manageable. Let's define the noise. Your phone's storage is filled with cached application data. Massive operating system updates, 4K movies downloaded for a flight, gigabytes of podcasts you've already listened to, and thousands of blurry, rapid -fire duplicate photos. None of that is your legacy. Exactly. When you isolate the pure human memories,

the curated photos, the specific videos of major life events, the scanned legal and historical documents, the scale shrinks dramatically. A comprehensive 2026 study of smartphone users analyzed the actual contents of median photo libraries. They found that the typical user has a library consisting of just 2 ,547 photos and 127 videos. That number feels shockingly low, but it really highlights how much of our perceived data bloat is just software overhead. And there

was another stat in that study from our sources that really proves the silent decay we talked about. Of the oldest photos sitting on a typical smartphone is only 2 .7 years. That is a staggering metric. It is. It proves that users are routinely, quietly losing their older photographic history. When they upgrade phones or when a cloud sync fails silently in the background, they lose the older files and simply adapt to the new baseline. They don't even realize the history has been

truncated. But let's look at the gigabyte footprint of those 2 ,500 photos because this is where the anxiety truly dissolves. Let's do the exact math on the camera roll. How big is a memory? A modern smartphone camera, typically 12 to 48 megapixels, uses highly efficient compression algorithms. Apple, for instance, uses the HEIC format, which is based on high -efficiency video coding. It utilizes complex spatial prediction to compress image data without losing visual fidelity. A standard 12 -megapixel HEIC photo

consumes roughly 2 to 3 megabytes of storage space. If you shoot in older JPEG formats, it might be 4 to 6 megabytes. Now, let's introduce the baseline archival medium, a standard single -layer optical Blu -ray disc, which holds 25 gigabytes of data. 25 ,000 megabytes. So if we do the division. One single 25 gigabyte disc can hold approximately 7 ,500 HEIC format photographs? Yes. A typical individual's entire smartphone

camera roll, representing years, potentially a decade of their life, fits onto one single physical disc with gigabytes of room to spare. Even for an avid amateur photographer who curates a library of 15 ,000 vital photos, their entire legacy is safely stored on two thin discs. It is the digital equivalent of a shoebox. You do not need an enterprise server rack. A physical Blu -ray disc is a millimeter thick. You can hold your entire photographic history between

your thumb and forefinger. Now, we do have to address the elephant in the room regarding storage space, which is video. Because modern phones default to shooting video in incredible resolutions. And video consumes storage at an exponentially faster rate than photography. Video is the hidden tax on storage, definitely. While videos might only represent 5 % of the files in your library, they will easily consume 40 to 50 % of your storage capacity. Let's look at the bit rates. One minute

of 4K resolution video, shot at a standard 60 frames per second, requires roughly 400 megabytes of storage. So you're eating up nearly half a gigabyte every 60 seconds you leave the camera rolling. Correct. Extrapolating that... An hour of standard 4K home video will effectively fill one 25 -gigabyte disk. However, even when factoring in heavy video usage, the scale remains remarkably finite. If we analyze a family of four, all taking photos and recording major milestones over an

entire 25 -year generational span, the mathematics show that a stack of just 20 disks is sufficient to archive the entire family's visual legacy. 20 disks. That is a stack roughly an inch high. sitting on a bookshelf. That is the true scale of your digital life. The risk isn't that you lack the capacity or the wealth to build an archive. The risk is simply procrastination. It is failing to spend the afternoon organizing those files and committing them to the shelf before the cloud

provider changes their terms or the hard drive seizes up. But this naturally leads to the most critical technical question of the entire deep dive. If you're nodding along, realizing you only need to burn 20 discs to save your family's history, you are also likely remembering the early 2000s. You're remembering buying spindles of cheap CD -Rs from an office supply store, burning a mix CD for a road trip, and finding that it skipped, failed, and became completely unreadable six months later. Why should anyone trust their family's legacy to a plastic disc?

What makes modern archival discs any different from the junk we used to burn? To understand the solution, we have to deeply examine the chemistry of permanence. We have to dissect why those old CD -Rs and DVD -Rs failed so miserably. A standard writable optical disc consists of a polycarbonate plastic substrate, a recording layer, a reflective metallic layer, and a protective lacquer coating. The fatal flaw in consumer CD -Rs was the composition of the recording layer. They relied on organic dyes. typically cyanine, thalassianine, or azo

dyes. When you say organic dye, you essentially mean complex chemical paints? Exactly. The burning process for a standard CDR involved the drive's laser firing into that organic dye layer, rapidly heating it, and causing a chemical reaction that permanently darkened microscopic spots. These dark spots and the unaltered clear spots represented the digital ones and zeros, the pits and lands, that a reading laser would later interpret. Right. But because the dye was organic, it was fundamentally unstable. The molecular bonds within the dye

were highly susceptible to environmental degradation. Ultraviolet light from the sun would literally break the double bonds in the dye molecules, causing the darkened spots to fade until the laser could no longer distinguish a 1 from a 0. High humidity would cause the layers to delaminate. Furthermore, if the laser in the disk drive was slightly weak, it wouldn't heat the dye sufficiently, creating a shallow burn that degraded even faster. Spindles of those early 2000s disks that appear perfectly pristine to the naked eye today will almost universally return massive read errors

on their inner tracks because the chemical dye has simply aged out of existence. So organic chemistry was the vulnerability. Enter the technological hero of this deep dive, the M disk. What exactly is an M disk and how does it solve the chemistry problem? MDISC, which stands for Millennial Disk, was a proprietary technology born out of extensive materials science research at Brigham Young University and commercialized around 2009. The revolutionary leap of the MDISC was the complete elimination of the unstable organic dye layer. Instead, the

MDISC utilizes a recording layer composed of an inorganic, glassy carbon material. Glassy carbon. So instead of a chemical paint, it is essentially a micro -thin layer of synthetic rock. Functionally, yes. It is an inorganic compound that is highly inert to oxidation, immune to ultraviolet degradation, and possesses an incredibly high melting point. When an M -disc is placed in a drive to be written, the recording laser has to fire with significantly higher power than it would for a die disc. It doesn't just induce

a chemical color change. The laser physically melts localized pits into that hard, rock -like carbon layer. The data is written via a permanent physical phase change in an inorganic material. That is fascinating. Yeah. You're essentially engraving data into stone at a microscopic level. And this isn't just theoretical marketing. There is a legendary independent test detailed in the source material that conclusively proves just how resilient this inorganic layer is. Let's talk about the China Lake Test, because the parameters of this experiment sound like a military torture

chamber for electronics. The China Lake Test is the foundational proof of concept for the MDISC. The United States Naval Air Warfare Center Weapons Division, located at China Lake, California, routinely conducts extreme environmental testing for military hardware. They decided to evaluate the archival integrity of the early M -Disc DVDs against the best, highest quality standard organic die discs on the market using premium brands like Verbatim, Mitsubishi and Delcon devices. OK, so top tier consumer stuff. Right. They placed

all the disks into environmental testing chambers designed to simulate severe accelerated aging. The disks were subjected to grueling temperature cycles reaching 85 degrees Celsius, 185 degrees Fahrenheit, combined with 85 % relative humidity, while simultaneously being bombarded with full -spectrum light arrays designed to mimic intense direct solar radiation. They basically threw absolute hell at these disks to see which one would break first. The results were stark and definitive. Every single standard organic dye

disk suffered catastrophic data loss. The environmental stress caused the chemical dyes to degrade and the layers to physically delaminate. The M disks, however, survived the entire testing protocol without any measurable data degradation. The Navy's report concluded that the M disk was the only format capable of surviving the rigorous environmental conditions. That test established the MDISC as the gold standard for institutional and personal archiving. OK, the chemistry is brilliant and the military test is highly convincing.

But I am going to push back here because I'm looking at the corporate history in our source notes. It clearly states that Millenniata, the company that originally invented and patented the MDISC technology, went bankrupt and dissolved in December 2016. Okay, how can we sit here and confidently advise the listener to entrust their irreplaceable family legacy to a format created by a defunct company? Doesn't a corporate bankruptcy trigger the exact administrative clock warning we spent the first 20 minutes of this deep dive

criticizing? It is a vital question, and it is a nuance that trips up many people looking into archiving. But the dissolution of Millennata actually represents a massive, necessary victory for the MDISC format. It demonstrates what archivists consider the correct order of operations for a sustainable medium. How does the creator going bankrupt represent a victory for the product? You have to separate the corporate entity from the open technological standard. Millenniatas struggled financially not because the M -Disk

technology was flawed, but because the broader consumer hardware market experienced a massive shift. Around 2015, manufacturers like Apple and Dell completely stopped including built -in optical drives in their laptops in favor of making devices thinner and relying on the cloud. The overall volume of optical disks sold plummeted. The startup couldn't survive the hardware transition. Oh, that makes sense. The market changed. However, the medium itself, the M -Disc standard, was so robust that it survived the death of its creator. Before millennia unfolded, they successfully

licensed the manufacturing technology to massive, diversified global conglomerates, most notably Verbatim, which operates under the umbrella of the Mitsubishi Chemical Group. Mitsubishi has the deep pockets and manufacturing infrastructure to keep producing the discs indefinitely, regardless of the fate of the original Utah startup. I see. The intellectual property outlived the startup, and manufacturing passed to a global titan. But what about the hardware to read it? If the company is dead, how do we know the software or the drives

will even exist in 20 years? This is the quiet, ultimate triumph of the M -Disk Blu -ray standard. Absolute read compatibility. To read a modern M -Disc Blu -ray, you do not need a proprietary specialized M -Disc reader. You do not need to download legacy driver software from a defunct 2016 website. You do not need a cloud API key, and you are not beholden to any corporate terms of service. An M -Disc Blu -ray is designed to meet the exact optical specifications of the standard Blu -ray format. It mounts seamlessly

in any cheap mass market $30 external Blu -ray drive you can buy off the shelf today using the standard 405 nanometer blue laser. It mounts natively in Windows, Mac OS, and Linux without requiring third -party software. That is the ultimate relief. We avoid the 1970s tape deck problem entirely. We have a universal... open standard. If I pull an M disk off a dusty shelf in the year 2045, I can simply plug whatever generic optical drive exists at that time into a computer, and the standard file system will mount the data instantly. The data is completely

sovereign. So we have established the perfect inorganic disk. The chemistry is sound, the capacity handles our camera rolls, and the standard is immune to corporate death. But building a true archive is not just about logging onto Amazon and buying a spindle of blank disks. It is about the execution. The physical act of archiving requires a specific protocol. Let's get incredibly practical. How do you actually build this archive this weekend? What is the architect's practice?

The practice of archiving requires a disciplined sequence. It is an intentional, deliberate process that is slightly dull, which is exactly why it is effective. The methodology breaks down into four essential steps. Freeze, hash, burn, and verify. Okay, walk us through them. First, you must freeze the folder. You gather the specific photographs, the finalized tax documents, the exported video files you intend to archive, and you place them into a single, dedicated directory.

And then crucially, you stop modifying them. You do not open them to make a quick crop. You do not rename them. the data state must be completely frozen. Because if you change a file after the next step, the entire verification process breaks. So what is the next step? Step two is creating a manifest of hashes. A digital hash is a fundamental concept in cryptography and data integrity. Using a standard algorithm like SHA -256, you use a simple software utility to scan every file in

your frozen folder. The algorithm processes the raw binary data of the file and outputs a unique fixed -length string of characters. This is the file's mathematical fingerprint. And the architecture of a hash is incredibly sensitive, right? Like, if I take a high -resolution photograph and change exactly one single pixel from dark blue to slightly darker blue, the resulting hash string will be completely, radically different from the original hash. That is called the avalanche effect. Even

a single flipped bit entirely alters the hash. By generating a text file containing the hashes of every photo you are about to burn, you create an infallible ledger of the data's exact state at the moment of archiving. Brilliant. Step three is the burn itself. You load your blank M disk and initiate the burn software. The critical rule here is to prioritize accuracy over speed. Modern optical drives boast incredibly high maximum burn speeds, 16x or 24x, which are essentially party tricks. Riding to that glassy carbon layer

requires sustained, deep laser power. If your software offers the option, you manually restrict the burn speed to match the baseline rating of the disk, usually 4x or 6x. A slow, steady burn using constant linear velocity ensures the pits are melted deeply and cleanly. And then comes step four, which is arguably the most important rule of this entire deep dive. Verification. Let me bring up an analogy. Going through the trouble of buying an M -desk, burning your files to it, and immediately tossing it into a desk

drawer without verifying the integrity of the burn is exactly like packing a parachute with a blindfold on, shoving it into a closet, and just confidently assuming it will successfully deploy when you jump out of an airplane ten years later. You have absolutely no idea if the burn process actually worked until you test it. The parachute analogy is perfect. A silent error could have occurred during the write process, a speck of dust, a micro fluctuation in laser power, a buffer underrun. To verify, you simply

instruct your software to read the freshly burned physical disk and recalculate the SHA -256 hashes of the files directly from the optical media. The software then compares those new hashes against the original manifest you created in step two. If every single hash matches perfectly, you haven't just copied data. You have mathematically proven, down to the exact binary digit, that your archive is a perfect clone of your original files. If the hashes don't match, you throw the disk in

the trash and try again. Blank disks are cheap. Corrupted memories are forever. Now, to ensure that the burn process goes perfectly on the first try, We need to talk about the physical hardware doing the burning. Because the source material provides very specific drive recommendations to avoid a mechanical issue known as starving the laser. What does it mean to starve a laser during a burn? This comes back to the physical demands of the M -Disc. Melting pits into inorganic glassy carbon requires significantly more thermal

energy than tinting an organic dye. The laser diode in the drive must maintain a very high, entirely uninterrupted power draw throughout the entire burn process. This is particularly crucial if you are writing to high -capacity BDXL disks. A standard Blu -ray holds 25 gigabytes, but a triple -layer BDXL disk holds 100 gigabytes. Burning 100 gigabytes at a slow, reliable speed can take over an hour of continuous high intensity laser firing. And this is where the power source becomes the bottleneck. Exactly. Many modern

external optical drives are bus powered, meaning they draw their electrical current entirely through the USB cable connected to your laptop. A standard USB port has strict power delivery limits. If you are halfway through a 100 gigabyte burn and your laptop screen decides to go to sleep or the operating system temporarily diverts power on the USB bus to prioritize another component, The voltage supplied to the optical drive dips for a fraction of a second. The laser starves. The thermal energy drops. The pit isn't melted deeply enough. The burn silently fails, and you

have just ruined a $30 blank BDXL disk. So how do we mitigate that risk? What specific hardware should the listener purchase to build their architecture? The recommendations are split into two tiers based on your capacity needs. The entry -level recommendation is for users who are primarily archiving documents and a moderate amount of photos onto standard 25GB M .Discs. For this, the Buffalo 6X portable Blu -ray drive is highly reliable. It explicitly states firmware support for the M .Disc standard. However, because it

is a bus -powered drive, you must follow strict operational hygiene. You must plug the drive directly into a primary USB port on the computer chassis, never into an unpowered USB hub, a monitor pass -through, or a cheap dongle. And crucially, you must configure your operating system settings to prevent the computer from going to sleep or spinning down the hard drives while the burn is active. You essentially have to babysit the laptop to ensure it keeps feeding power to the USB port. But what about the Pro tier? For the person shooting 4K video, who needs to utilize

the massive 100 gigabyte triple layer M disks? For massive hour -long burns, you want to eliminate the USB power variable entirely. The top line recommendation is a desktop class drive, specifically the Asus BW16D1X -U. The model number is obtuse, but the engineering is what matters. The Asus drive does not rely on your laptop's USB port for power. It features its own dedicated power brick that plugs directly into an AC wall outlet, pulling a highly stable 12 volts at 3 amps. This

means the spindle motor turning the heavy 100 gigabyte disk and the 405 nanometer laser melting the carbon layer have access to massive uninterrupted current directly from the wall. The laser never begs the laptop for power. It delivers incredibly stable, deep burns. So, we have the correct high -power Asus drive. We've executed our slow burn and we verified our hashes. The data is mathematically perfect and permanently etched into synthetic rock. Now, what are the physical storage rules?

Because the physical treatment of the media dictates its longevity. I remember back in the early 2000s, people would buy those sticky printable paper labels and stick them onto the top of their CDRs to make them look professional. Is that acceptable for an archive? Under no circumstances should you ever apply an adhesive paper label to an archival optical disc. At the high rotational speeds required to read a Blu -ray disc, a paper label that is even a fraction of a millimeter off -center will introduce a severe physical

imbalance, causing the drive to vibrate violently, which can induce read errors or physically shatter the disc inside the drive. Furthermore, the chemical adhesives used on the labels can eventually off -gas and degrade the protective lacquer layer on the top of the disc. Write on the inner plastic hub with a specialized soft -tip marker if you must, but ideally, keep the disc face pristine. And how should they be housed? Store the burned M discs in standard, rigid plastic jewel cases. Never use those cheap, thin paper sleeves or

flexible plastic binders. Every time you slide a disc in and out of a paper sleeve, microscopic dust particles act as an abrasive, eventually scratching the polycarbonate data surface. Store the jewel cases upright, exactly like books on a library shelf. Do not stack them flat in massive towers as the weight can cause the bottom discs to warp over decades. Keep them in a cool, dark, and dry environment. An interior closet in a climate -controlled living space is ideal. Do not store them in an uninsulated attic where wild temperature swings will physically expand

and contract the polycarbonate layers, inducing stress fractures. And certainly do not store them in a basement where a flooded water heater could submerge them. And the labeling on the case is vital. As the old archivist saying goes, a perfectly preserved disk that nobody can identify in the year 2045 is only slightly more useful than a disk nobody can read. Print out a paper insert for the jewel case, detailing exactly what is on the disk, the date it was burned, and what specific hashing algorithm you used

to verify it. Which brings us to the ultimate architectural rule of thumb for this entire practice. The rule that guarantees survival against physical reality. Right. It is a variation of the classic 3 -2 -1 backup strategy, adapted for physical media. The rule is this. One single disk is merely a hope. Two identically burned and verified disks stored in two different rooms of your house is a reliable method. Three disks with the third stored entirely off -site, perhaps in a bank

safety deposit box or simply on a shelf at a trusted friend or family member's house in a different city, is the standard you use for the human records you can absolutely never afford to lose. A localized fire or flood can destroy two disks in one house. It is... statistically impossible for two separate natural disasters to simultaneously destroy the primary home and the off -site location hundreds of miles away. We have covered immense ground today. We have examined the silent, relentless threat of the

administrative clock and the physical decay of magnetic media. We have shattered the illusion of data massiveness, reducing decades of human memory down to the scale of a single physical shoebox. We have explored the glassy carbon chemistry of the M disk that solves the degradation problem. And we have laid out the precise hardware, the cryptographic hashing, and the physical practice required to execute this successfully. Let's bring this all home and summarize the core practical value for you, the listener. Against a decade

of paying monthly cloud rent to a tech conglomerate, Against the constant low -level anxiety of getting locked out of an automated account or having your data held hostage by egress fees? A stack of verified, verbatim MDISCs and a dedicated ASUS optical drive represents a financial rounding error. We are talking about an investment of less than $150. But what that small investment buys you is functionally priceless. It grants you true, API -free, terms -of -service -free, absolute custody of your life's irreplaceable

memories. No tech billionaire can unilaterally change the terms of a physical M disk sitting quietly on your bookshelf. No server migration can erase it. It is entirely sovereign. It is the purest form of digital sovereignty available to a consumer today. But as we conclude this analysis, I want to leave everyone with a broader, deeply philosophical thought to mull over. It builds on everything we've discussed today regarding the fragility of the web. the great divergence, and the illusion of permanence. Historians and

data archivists recognize that we are currently living through a profound paradox that is increasingly being referred to as the digital dark age. A dark age? How is that possible when we just established that human civilization is producing 600 exabytes of high -fidelity data every single day? We are drowning in information. We are drowning in information, but we are starving for permanence. We produce exponentially more raw data daily than any previous civilization in human history. Yet precisely

because of the ephemeral nature of magnetic charges on SSDs, the seizing fluid dynamic bearings of hard drives, and the incredibly fragile rental agreements of the cloud, we actually run the very real risk of leaving behind fewer permanent readable records than the ancient Sumerians did. That is a staggering. realization. The Sumerians inscribed cuneiform onto clay tablets. Whether they intentionally baked those tablets in kilns or whether the tablets were accidentally fired when their cities were burned by invaders, that

inorganic clay survived in the earth for 5 ,000 years. We can read their grain receipts and their poetry today. Exactly. A fired clay tablet is essentially a primitive macroscopic version of the glassy carbon layer of an M -disc. It is a physical phase change in an inorganic medium. So the provocative question we must ask ourselves is this. If a future digital archaeologist, or on a much more personal and realistic timeline, simply your own great -grandchildren 50 years from now, decides to dig through the remnants

of our current era to understand who you were, what exactly will they find? Will they find your carefully curated, mathematically verified M -discs, resting safely in a labeled jewel case on a bookshelf, ready to mount in a standard drive? Or will they simply uncover the rusted, completely unreadable platters of an old mechanical hard drive sitting at the bottom of a flooded desk drawer, alongside a dead web link to a cloud account that a corporation automatically deleted decades ago? It is completely up to you to leave a deliberate fossil record. You do not have to

save the noise. You don't need to archive the cached application data or the downloaded movies. But the memories, the photographs, the written documents, the audio recordings that prove you were here. and that you lived a beautiful, complex life. Do not leave them sitting on death row in a corporate server farm. Take the afternoon. Freeze the folder. Generate the hashes. Burn the inorganic copy. Verify the math. Put it on a physical shelf. The cloud can comfortably hold your working set while you need immediate access

to it on your phone, but the shelf holds the incontrovertible fact that you had a life worth saving. Take back absolute custody of your data. We will see you on the next deep dive. Welcome to the multiplex where the future comes alive. Past, present, future tech. Open up your eyes. Brian Romella leads the way with research deep and true. Technology and history, he connects

them all for you. Brian Romer leads the way with research deep in true denology and history. He connects them all for you. Ancient sparks to modern code. He traces every thread. Go exploration. Knowledge widely spread. Read multiplex. Go come.

But humanity moves on. Love equation for AI. Zero human companies build. Thermodynamic wages earning the key. Secrets we will read, experiment in publishing, real research and expanding every great complex.

Open the transcript file

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. ReadMultiplex.com: The Field That Heals: Why Pulsed Electromagnetic Fields May Have Profound Benefits.

    Nov 21, 202651 minAudio

    Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Every cell in your body has lived its whole life inside a magnetic field. The Earth is a dynamo, and it has been...

    Credit the people on this episode

  2. 78 RPM Record Hour: From Steel to Hillbilly

    Nov 19, 20261 hrAudio

    Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 3: From Steel to Hillbilly (Early Country). From steel to hillbilly: Frank Ferera,...

    Credit the people on this episode

  3. 78 RPM Record Hour: The Knife and the Steel

    Nov 19, 20261 hrAudio

    Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 2: The Knife and the Steel (Hawaiian). The knife of the blues slide guitar and the...

    Credit the people on this episode

  4. ReadMultiplex.com: How to Run Local AI on a Computer From 2017.

    Nov 18, 202635 minAudio

    Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Somewhere in your house there is a computer you stopped thinking about. Maybe it is a 2017 laptop in a drawer, its...

    Credit the people on this episode

  5. 78 RPM Record Hour: Blues (Pilot)

    Nov 18, 20261 hrAudio

    Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 1 (Pilot): Blues. Fourteen original 78 RPM records from 1917 to 1925 that trace...

    Credit the people on this episode