On the Physicality of Data
In some ways, this tidal wave of data is the pollution problem of the information age. All information processes produce it. If we ignore the problem, it will stay around forever. And the only way to successfully deal with it is to pass laws regulating its generation, use and eventual disposal.
– Bruce Schneier, “The Future of Privacy”
Data has a physical presence.
We think of data as ephemeral. It flies unseen through the air and pops up, often unbidden, on devices. It slogs through cables between machines. It flies as flashes of light between satellites or through distant underwater cables. And then deleted. Forgotten. A data point never seen or shared again. Data is knowledge, shared among all humankind they say. Or, perhaps, shared between indentured employees dwelling deep within these highly classified bowels of paranoid, highly militarized, crypto-corporate overlords.
But in reality, data* is an actual thing that takes up real space. An electron has mass.
Hold up an old, conventional solid state drive. Open it up with a screwdriver. Inside it’s nothing but line after line of semiconductor chips holding non-volatile memory. A chip is a couple of ounces. That entire box weighs a few pounds. That few pounds holds one trillion – 1,099,511,627,776 bytes. 2^40 bytes or 16^40 bits – ON or OFF. 8 ons or offs makes a byte, a series of bytes makes data.
Consider the lowly bit. The bit is the cosmic pile-up of electrons somewhere deep in the strata of the semiconductor. Enough electrons generates the voltage and tells the system yes, there’s data here. Too little and it’s a no. Flip a bit from 1 to 0 and the electrons all vacate and move to some more interesting semiconductor. This is not nothing.
With the complexity and VR simulation of the modern decking experience, today’s decker rocks disks 10-20 times that size. Some of the newest, hottest bit of kit rock solid state drives up to 100T. With advancements in semiconductor technology, again from the paranoid, crypto-corporate overlords but our crypto-corporate overlords, the decker’s kit hardly weights more than a few pounds for those electrons all packed in an itty bitty space.
But the modern paranoid, highly militarized, crypto-corporate overlord isn’t satisfied with a few Ts of storage. Every bit of their data – every bit – generated by their employees or their users goes to some enormous back-end to feed the maw of a hungry processing engine to send analytics back into the corporation. From analytics, they try to figure out:
- How do users consume their product?
- What products do consumers buy and which do they ignore?
- What are the measurable metrics around employee performance? Are we measuring them every minute? Every second?
- Are employees generating non-sanctioned data on work hours by consuming competitor product?
- Have we conquered the world with our evil yet? Is that a measurable metric?
From what looks like a bunch of random records, the corporation makes more data from its data and feeds that data back into its data to make more data. And this data piles up. They build enormous data centers underground. The disks require more and more space. The ravenous maw of silicon forces more silicate substrate, more drives, more data.
Moving it is a real pain for the corporation. But it’s an opportunity for shadowrunners.
Moving Data
The evil crypto-corporation blew out their current datacenter’s capacity. They’re out of disk space and physical racks to build more. They built a new datacenter out underground in the desert with state of the art physical security – enormous armed drones, guns, satellite surveillance, smart cameras, walls, barbed wire, and smart-wired in security in battle suits.
The new datacenter is enormous. It can hold a Yottabyte. For reference, a Y is 1000 Zettabytes, and a Zettabyte is one million of those decker hard drives lashed to that deck. To build this monstrosity, the corporation conquered, aggressively colonized, and mined for silicon some third world Asian country to get those semiconductors made on the cheap. But hey, that was a third world country, and not here in Seattle, so who cares about them. Back to the here and now.
Problem is, the corporation has to move the data from its old datacenter here, on the sixth floor of their current corporate arcology, to the new hyper-cooled datacenter under the desert out in what was Nevada back when it was the good old US of A and was the Ute Nation but today is now the Pueblo Corporate Council. And data, as alluded to above, is ultimately made of electrons sitting in microscopic doping buckets in semiconductors. Neither those electrons nor their semiconductors are ephemeral. Equipment doesn’t merely poof from one datacenter to another.
One would think a sizeable corporation could pay for laying cable between their Seattle corporate arcology headquarters to their Pueblo Corporate Council site. And they’ve done precisely that. They negotiated treaties, one side got screwed, lawyers argued about the intellectual property integrity of the data, and money exchanged hands.
Even with the hyper-fast cable, the electrons must convert to light and get transmitted down the cable. And even at speed of light transmission through optical cables, the cable has limitations on how many electrons it can encode and decode from light back to data. That all takes time. By calculations of the corporation’s best engineers, their new datacenter will populate with copies of the current data and backups in about seven and a half years.
The corporation cannot survive without its data from its data for seven and a half years. How will it make decisions? How will it figure out what nations to destroy? What evil products to work on? Or when to put its money in unspeakable gene-splicing experiments in the service of Insect Gods and their Shamans?
And what happens if they split brain their data, with old data in one location and new data elsewhere? How will they run their analytics?
There’s an easy and obvious solution to the corporation’s problem:
Physically load the data on trucks and move it from Seattle to the datacenter under the Pueblo Corporate Council.
It turns out, the bandwidth of sneakernet is vastly higher than the bandwidth of a fiber optic cable. Even in the far-off cyberpunk future.
The Job
Usually a job goes down like this:
- Someone meets up with a Mr. Johnson in a bar.
- The Mr. Johnson wants some useful data-slash-dirt on someone.
- The data belongs to a paranoid crypto-corporation that is quite clearly evil/amoral/full of dudes who need killing.
- Shadowrunners sign up!
- Riggers do awesome things with drones.
- Street sams fire Ingram smart guns.
- A decker, who was sitting in the back watching old Star Wars movies off her deck while the job was going down, jacks in.
- The decker has some cool cinematic fights, largely in her head, with Black IC. And then she scores a single file about a thing.
- The Shadowrunners get out.
- Mr. Johnson ponies up for the file. Perhaps he has another mission for them. Perhaps not.
- Profit!
Maybe during the course of a run a few shadowrunners roll bodies and take stuff. But, no one is really hurt in the excitement except for armed guards (monitored via data every second of their work day) who protected access to that single file in the first place. And they’re basically human packing peanuts in the grand scheme of the universe. The real goal is whomever was in that file.
But this time it’s different.
It’s impossible to put a dollar value on the worth of a refrigerated semi carrying the data of an entire corporation. One file is valuable, sure. Maybe a few files are valuable. But an entire truck worth of files…
They call it Project **Snowball.
The Shadowrunners cannot hit the data while it’s in ensconced in a corporate datacenter. If Shadowrunners thought breaking into a corporate office to steal a single file was bad, they have seen nothing yet. The datacenters are, quite literally, designed to take direct hits from a nuclear, or dragon-based, exchange. Even the datacenters in the local arcologies are physically impenetrable. The heaviest of heavy weaponry isn’t getting through the doors. No way, chummer. No way.
But a semi is only as good as the defenses in and around the semi. To a good Shadowrunner, it’s a soft, squishy, data-filled marshmallow.
For a couple of hours, an entire crypto-corporation is exposed.
The semi will look like any old innocuous but enormous refrigerated grocery truck. It will bear no markings of the corporation. The corporation will periodically switch drivers along the route so no one driver knows the entire route nor what they’re hauling. Corporate security will arm the truck with the very best in automated drone-based guns. Lots of guns.
When the truck leaves the originating facility is highly classified (see again: data). The route is also highly classified.
Hitting the truck while the corporation loads the data is no good. The corporation will defend the data and the truck within an inch of their lives. And hitting the truck at its destination is no good either. It’s essentially in an underground militarized zone in a hostile (to Shadowrunners) nation.
Hitting the truck while in motion on the road will draw tons of attention, the cops, the news, and the corporate goons. It’s possible but non-optimal. It’s a failover plan.
The corporation’s paranoia is its weak point. Hit the truck while changing drivers at trucking rest stops.
Here’s what the Mr. Johnson says about the run:
Chummers, I have a mole on the inside who will tip you off when the semi leaves the facility fully loaded. They will give you intelligence on its license plate and markings. We don’t know the driver change points but you can tail it and use drones for aerial surveillance. Get the semi while the drivers change and stop for coffee. Hit them hard. And get me that semi. My employer can finally take down the evil, paranoid, crypto-corporation. You want to do good, right? They’re evil and we must stop them. The payout will ensure you can retire from this life.
And the payout is enormous. Absolutely enormous.
The Mr. Johnson provides an address in what was Utah in the Ute Nation. On a map, it’s simply another rest stop. Once there with the semi, the Shadowrunners will hand over keys to their contact, and the mysterious backer will hand over suitcases of cash. They will shake hands and walk away. The backers will provide a car to return the Shadowrunners to Seattle.
This sounds like a prime run. For Shadowrunners, this is a couple of exciting, high-speed chases as someone mysterious tries to wrest the semi from their hands. Preferably in some dark tunnel. The rigger will get to use all her drones and vehicle-mounted guns. The street samurai will drive a semi while attacked by vehicles on all sides like Furiosa in Fury Road. Someone, preferably a corporate goon, will die shiny and chrome. And it’s a huge potential payout. Huge.
Hold up. Logically, the decker points out, the only people who can house, utilize and process this much data are:
- Other enormous rival crypto-corporations who already have their own desert datacenter and want the copy to take down their rival;
- State-actors who want the data for their dark intelligence operations;
- Someone unknown and truly horrible who has the processing capacity to take the data and do something interesting with it.
Normal people can’t use the data simply because they have no physical room to house it all. It’s too much. It’s too big. Even getting the processing streams to analyze the data requires specialized skills and enormous systems.
And worst, the original corporation is temporarily inconvenienced from a data standpoint. It’s not like they deleted their data. They made a copy, and simply doubled their electrons and silicon doping on that semi. It’s the facts the rival learns through analyzing the corporation’s data that worries them. Just a huge dump of data is worthless to most people. But to some, very specific, people…
This is like disposing of toxic waste except a million times worse. At least no one wants the toxic waste. Everyone wants a semi full of highly classified corporate data.
Whoever is losing the data is going to want to get it back.
Whoever is getting the data will desire it at all costs.
It may be worth more, in the long run, for the Shadowrunners to keep the data instead of taking the payout.
And the Shadowrunners can drive it anywhere, including to a new, higher bidder, as long as they can find working diesel stations along the route. As long as no one hits the semi with a huge electro-magnetic surge to reset all the electrons, or otherwise damages the equipment, the data never rots, never goes stale, never loses its worth. Even ten years from now, that much data is incredibly damaging to the host corporation. Something is in there. Lots of somethings. Something always is.
What the Shadowrunners do with data from the shootout at the donut and coffee rest station on the side of the highway is up to them. They can keep it. They can hand it over for their payout. Or they can sell it to a higher bidder.
But what they are, is now saddled with a corporation’s worth of data on wheels.
And it’s enormous and requires diesel to keep it moving.
* Data is technically plural. The singular is datum but in our crypto-corporate future, the PR marketing departments have carefully manipulated language through aggressive marketing launches and now we use the singular with a plural noun.
** After Amazon’s Snowball, of course.
Image Credit: Art by Jaydot Sloane of Vanity Games – http://www.patreon.com/VanityGames
robbbbbb says
If I’m the evil corporation, I put a dead man’s switch, or a remote trigger on that can wipe all the data if it gets seized in transit. Press a button, and the whole semi melts down. (Metaphorically; it all just gets erased. I mean, yeah, you could have a really large charge of thermite suspended above the data drive and a rather spectacular meltdown, but I digress.)
So if the Shadowrunners want to do something more than lightly inconvenience the major corporation, then they’re going to need to do something about that dead man’s switch.
Daniel Carrier says
I’d just encrypt it.
SunwolfNC says
Remote triggering of an EMP device embedded in the truck container, yep. Like Rob said.
I’d almost expect something like that if I stopped to clear the Nuyen symbols from my eyes in time…
There are so many angles to a heist like this that’d have me tied up in knots trying to plan for it.
Anyone with that much data will protect it. But how? Satellite monitoring and vectoring in company men to stop the heist? Hacking street cams to follow the progress and then plan for an interdiction of the truck? Beacons embedded in the vehicle that phone home the second the ‘I’m OK’ button isn’t pressed every 18 minutes?
Shotgunning missions this big is what got my Face-man/Rigger strung up in a warehouse…never to be seen again…*sniffle* miss that guy…
After that episode, I took it upon myself to be extra detailed in scoping jobs for my team. I scanned everything, and everyone. For anything. Ran background checks, tailed Johnson’s around to verify their story, and so on. Ran into a couple of nice side story/missions doing it, too 🙂
When will someone make a faithful-to-pencil-n-paper-Shadowrun MMO? I’m still yearning for one!
Eran says
As someone who works in data migration, you’re not far off, but migration via truck isn’t used. Data is migrated system-by-system, and enough bandwidth is usually available to do so within months. Even if it would take years, it’s not a problem because each system only takes a small fraction of that.
Ben McFarland says
I’m much more boring. I mirror the zfs drives over VPN and we duplicate the volumes while never taking downtime, and when the mirrors are silvered, we split the pools, remirror with fresh hard ware and it’s done without ever leaving the data center. ZFS and Solaris are awesome. We just have [REDACTED] PR. I can do this right now, I don’t need the future, and I don’t need a semi.
Ben McFarland says
Although, in want of a good scenario, you instead get smuggled into the new secondary data center, or logistics center, and you plant a Trojan drive under the cover of a separate op, and you cause a component failure requiring hardware replacement. That op plants a rogue limited AI which then piggybacks the mirror process and transmits it to a drone relay which, in turn, routes the duplicated data to the third party who commissioned the OP. The final outcome remains compartmentalized and the runners lack the knowledge to double-cross Johnson. People get paid, everyone is happy.