Cyberpunk Fantasy Asymmetric Threat summary

[Chamomile]

The place the gazetteer of time delays plan really falls apart is that planets orbit, so the time delay wouldn't even be consistent. They wouldn't even be consistent at any given time of year on Earth, because different planets rotate at different speeds. You'd have to figure out the game's current date and cross-reference a chart that spanned that a decade or more.

[deaddmwalking]

If you're talking about locations other than Earth, maybe?

We know that light takes 8 minutes to reach us from the sun.

If you're talking about light-speed communication (such as radio transmissions) if Earth and the target (outer) planet are in-line with each other and the sun (same side) you would reduce the time of communication by 8 minutes. If the target planet were on the opposite side of the sun from the Earth (again in a straight line) you would add 8 minutes.

Generally speaking, that's not going to change the timeframe (ie, 35-51 minutes to Jupiter is well out of combat time).

With Venus, you're looking at 2.5 minutes to 8.5 minutes.

For the sake of convenience, you can take a base time (mean distance) and add/subtract to get a realistic number.

Planet Distance in AU Travel time
....................................................................
Mercury 0.387 193.0 seconds or 3.2 minutes
Venus 0.723 360.0 seconds or 6.0 minutes
Earth 1.000 499.0 seconds or 8.3 minutes
Mars 1.523 759.9 seconds or 12.6 minutes
Jupiter 5.203 2595.0 seconds or 43.2 minutes
Saturn 9.538 4759.0 seconds or 79.3 minutes
Uranus 19.819 9575.0 seconds or 159.6 minutes
Neptune 30.058 14998.0 seconds or 4.1 hours
Pluto 39.44 19680.0 seconds or 5.5 hours

[Username17]

Doesn't it though? I thought usually the answer to the question "why doesn't the max level Uberwizard at the end of the adventure path gank the party at level 5?" was usually "mmm hrm uh what's that over there?" Age of Worms certainly didn't give any reason for why the thousands of years old sorcerer with contact other plane and nothing better to do hadn't had the party's great grandparents assassinated decades ago. That's contact other plane and thus time travel communication, but even simple instantaneous travel and communication really screws up a lot with the types of sneaky hit and run tactics the side without the overwhelming force wants to use.

Essentially instant communication is the default of modern existence. You've never lived in a time when a verbal command would have to be relayed by runner or rider, and thus you've never experienced a time when it was reasonable for "real time" orders to have transit times measured in minutes or hours. That's totally outside your lived experience, and the declaration that anything works like that in a setting is just as much a strain on your suspension of disbelief as the inclusion of vampires or form changing robots. The fact that factually sending radio wave communications to Mars would work that way in the real world is beside the point - what you're asking is for people to accept a world where the hacker operates as if they were sending punch cards via pneumatic tube. And that is going to feel deeply strange.

Simply speaking: relativistic communication times are completely inappropriate for any setting that is supposed to "feel modern" because modern communications do not work that way. Relativistic communications are the providence of space as nautical metaphors. The slow turnaround can be the time of sending smoke signals or releasing messenger birds from one sailing ship to another. But unless you're doing the full Honor Harrington and just cribbing Hornblower books but in space, there's not really any place for that sort of thing in science fiction.

And it's totally not necessary. Stranger Things takes place in the eighties, where people have radios and landline telephones, and those communications are effectively instantaneous within that setting. This in no way renders it impossible or even difficult to arrange scenarios where characters don't get important messages in time or have response times to communications that are longer than needed. The travel time of the information itself is utterly unneeded for any form of isolation to be established. Within the modern experience, information travel time isn't really a thing, and all tension is set up with other frameworks than herald literally running to impart the missive.

The place the gazetteer of time delays plan really falls apart is that planets orbit, so the time delay wouldn't even be consistent. They wouldn't even be consistent at any given time of year on Earth, because different planets rotate at different speeds. You'd have to figure out the game's current date and cross-reference a chart that spanned that a decade or more.

This is another reason, yes. In single author fiction, you can arbitrarily declare whether Mars is at its closest approach to the Earth (4 light minutes) or its farthest (20 light minutes) at whatever crucial point radio waves have to go from Earth to Mars or vice versa or both. If you really care about accuracy you can look up a chart and figure out what days and years your story could take place in given the turnaround time you wanted for your messages and replies to make your plot work. In an RPG, you have multiple people who are contributing to the story and might plausibly care what the lightspeed travel time is. That means you have to keep it consistent, and that means very closely tracking the exact date and keeping accurate astronomical charts and literally doing three body physics problems.

It's levels of realism that you'd never even consider applying to any other part of the game, and it's puzzling that people would think they'd want to do it for something as weirdly specific as phone lag on Mars.

-Username17

[Foxwarrior]

Radios and landline telephones are way more responsive than modern means of communication
But if space is an ocean, and you don't want oceans in your game, why into space?

[deaddmwalking]

Radios and landline telephones are way more responsive than modern means of communication
But if space is an ocean, and you don't want oceans in your game, why into space?

There is a difference between real-time communication and instant travel.

When I call someone in Paris, I can communicate instantly, but we're talking several hours before we can connect physically with the technology we have today (planes).

In Star Wars, having the Emperor call Vader on the holophone and having a conversation advances the story, even if it would take days to bring the ships together.

If Space is an ocean, the question is whether it's an 18th century ocean where two ships can 'pass in the night' or if it's a 21st century ocean where every ship is in communication with every other ship and satellite tracked.

[maglag]

Essentially instant communication is the default of modern existence. You've never lived in a time when a verbal command would have to be relayed by runner or rider, and thus you've never experienced a time when it was reasonable for "real time" orders to have transit times measured in minutes or hours.

So, are you a real life wizard with relics? Is magic real after all?

Because my smartphone still needs to:
a) be pulled from storage place.
b) be unlocked.
c) select phone call operation.
d) select specific person to call or write a number.
e) wait for them to actually take the call.

And all that added up can and often does take several minutes.

And if the smartphone ran out of battery, or if it's in a place with no network, or the subscription was not charged, or any of those apply to the other person, or if they're simply not near their phone or left it to vibrate inside a bag, then tough luck, "hours for them to get the message" is back in the menu. And if it's something really urgent, somebody will have to physically walk/run to their place and directly speak to them. Riding vehicles to pull it off definitely happened no few times in my lifetime.

This is another reason, yes. In single author fiction, you can arbitrarily declare whether Mars is at its closest approach to the Earth (4 light minutes) or its farthest (20 light minutes) at whatever crucial point radio waves have to go from Earth to Mars or vice versa or both. If you really care about accuracy you can look up a chart and figure out what days and years your story could take place in given the turnaround time you wanted for your messages and replies to make your plot work. In an RPG, you have multiple people who are contributing to the story and might plausibly care what the lightspeed travel time is. That means you have to keep it consistent, and that means very closely tracking the exact date and keeping accurate astronomical charts and literally doing three body physics problems.

It's levels of realism that you'd never even consider applying to any other part of the game, and it's puzzling that people would think they'd want to do it for something as weirdly specific as phone lag on Mars.

There's this thing called "RNG" for when the physics become too complicated.

Like weather calculation is hard work, so instead you roll into a random chart to know what weather is it in the game.

So just make a random inter-planetary lag table and done, easy peasy.

[Thaluikhain]

And if it's my phone it might tell you you've got a text message a day or two after it was sent. And while the GPS is decent, it's always says I'm facing a totally wrong way.

More on-topic, though, you might have real-time communication, but that doesn't mean you have secure and discreet real-time communication. Could well want to keep radio/whatever silence.

How that makes a difference, though, beyond allowing for "deliver this message" as a quest, I can't say.

[Korwin]

I'm pretty shure magical communication is not fail proved in AT and can be jammed.
Also while the Information travels instantly, the sender will still need time to think of an answer before he can answer something.

As for the Telepath example earlier in the Thread. I would assume retransmitting an message takes some non zero time and is in danger of serios "Stille Post" syndrome.

Stille Post = the game where you sit in an line with 10 (or 5, or 12, or whatever) other People, the on sitting on the left wispers something to the one to his right, who wispers the same message to the one to his right, and after the message got the the last person he gets to reveal loud what message got to him. How is this game called in English?

[Thaluikhain]

How is this game called in English?

"Chinese whispers" usually, though I don't know what the connection with China is.

[angelfromanotherpin]

How is this game called in English?

"Chinese whispers" usually, though I don't know what the connection with China is.

That's the British English name. In the states it's called Telephone. The Chinese connection is unsurprisingly racist.

Historians trace Westerners' use of the word Chinese to denote "confusion" and "incomprehensibility" to the earliest contacts between Europeans and Chinese people in the 17th century, and attribute it to Europeans' inability to understand China's culture and worldview. Using the phrase "Chinese whispers" suggested a belief that the Chinese language itself is not understandable. Additionally Chinese people have historically been stereotyped by Westerners as secretive or inscrutable. The more fundamental metonymic use of the name of a foreign language to represent a broader class of situations involving foreign languages or difficulty of understanding a language is also captured in older idioms, such as "It's all Greek to me".

[Korwin]

Historians trace Westerners' use of the word Chinese to denote "confusion" and "incomprehensibility" to the earliest contacts between Europeans and Chinese people in the 17th century, and attribute it to Europeans' inability to understand China's culture and worldview. Using the phrase "Chinese whispers" suggested a belief that the Chinese language itself is not understandable. Additionally Chinese people have historically been stereotyped by Westerners as secretive or inscrutable. The more fundamental metonymic use of the name of a foreign language to represent a broader class of situations involving foreign languages or difficulty of understanding a language is also captured in older idioms, such as "It's all Greek to me".

Thats allready racist (against Chinese People)? If anything it reads like, Westerners are dumb to me

[OgreBattle]

Is the premise of "Software is dead" based on any real world trends? I really like the idea of nintendo cartridges for hacker stuff with the decker-deck aesthetic.

In a cyberpunk future 50 years from today, it's difficult to say what computing will really be like. Disposable optical CPUs whose input fibers are welded to microdot-sized firmware cartridges, DNA-based biological computing, and passive optical neural networks built out of glass can all be imagined today. What actually gets realized in a fictional setting can be pretty much whatever you want.

It's worth differentiating between writing software and defining the behavior of a computational device. (We call the latter "programming," and it is closely associated with software, but a future where software is too slow to be worth worrying about might repurpose that word.)

Whether you're designing specialized hardware or doing basic scientific research, you have to fall back to general-purpose tools when you're in new territory for which no off-the-shelf solution exists. It's so damn useful to be able to write an arbitrary program and have a general-purpose computer run it for you that I can't see "write program, run program, look at results, repeat until coffee break" going away. In a brave new world of deep neural networks, that loop looks like "specify network architecture, input data, and tuning parameters; run training loop; evaluate trained network." Even if software goes the way of the telegraph, programming isn't going anywhere.

Depending on the economics of computing platforms in Asymmetric Threat, the R&D programming loop might be realized as: "corp scientist edits program, one-off chip is fabbed, chip runs program, results are recorded, chip is thrown away." This has some desirable characteristics:

• It un-democratizes computing. This gives us more of a retro-80s vibe, and it is good for a dystopian corporate future. Moreover, totalitarian DRM is now much more feasible, and having even the simplest of tools that allow you to peek behind the curtain and jailbreak someone's commlink or Compute For The Masses really requires you to go out of your way and is easily criminalized. This is good for cyberpunk.
• Corporate computing resources are now a target to be raided. Instead of breaking into a supercomputer to steal cycles, you can Hack the Gibson by doing a run on a fab facility to make a few illegal chips that will allow you to crack the security of your real target.
• Hard-coding computing device behavior is probably very environmentally unfriendly. (Imagine future computing devices that are even more disposable than today's smartphones.) This is good for setting and tone.
• "Computer programs" are now physical tokens. They can be unique, they can be uncopyable (due to impossible-to-duplicate characteristics of their manufacture), you can permanently destroy them, and you can get into real-world shootouts over possession of them. These are all desirable characteristics of valuable equipment in a tabletop RPG.

It could be that "software" is alive and kicking in the world of the super-elite who have access to quantum mainframes. But that could be a plot point to be explored during a campaign; the reemergence of general-purpose computing for the masses (e.g., because someone finally figured out how to make micro-quantum computers) and the top-down opposition to it would be a fun way to play out some of the culture wars of the 80s and 90s.

Something I don’t have a good mental picture of is how much more powerful is the specialized chip vs general hardware running software. Say sapient AI being unique hardware brains is cool, but for things beneath that.

Then again I don’t know what ‘real hacking’ is like so saying this settings cyber combat requires specialized hardware, and general software is just not strong enough, works for me.

[pragma]

Something I don’t have a good mental picture of is how much more powerful is the specialized chip vs general hardware running software. Say sapient AI being unique hardware brains is cool, but for things beneath that.

Much more. An easy comparison is looking at power/performance of a GPU vs. a CPU on the same task. This blog post has an example showing a dedicated GPU beating a CPU by a factor of 15 in performance. This example quick and dirty, a little imprecise (eg ignores power) and is conservative. I've seen literature where specializers for narrower tasks can often outperform dedicated circuits by factors of 100-1000 under the same power budget.

Specializers are also becoming increasingly interesting as transistor density increases because the leakage current of idle cores and, especially, of the cache needed to keep them coherent is starting to drag on energy per operation gains. Heavily duty cycled specializers are a way to use additional silicon area & transistor count to improve performance without paying a steep penalty in dark current. The main barrier to wider adoption of specializers is the additional design effort required to implement them. It's easy to imagine software and manufacturing practices catching up with that problem in a cyberpunk dystopia.

A note on general purpose software: I need to second Czernobog's point that it's never going away. Even in analog chips like radios, it's common to have small amounts of embedded digital for configurability, and embedded, programmable digital signal procesing is ubiquitous. It's very easy to have small programmable cores embedding in almost anything now, so I don't see that going away in the future. That said, having them be so slow as to be irrelevant is a good setting conceit, and their presence gives a hacker some way to modify/interact with specialized chips that would otherwise have their functions fixed at the factory.

A note on quantum computers: they are basically the opposite of general purpose software, so the wealthy having access to them doesn't necessarily mean they get to program while everyone else has to fabricate. They mostly perform really well on simulating molecules and cracking codes, so miniaturized quantum computers are great examples of the exotic specializers a criminal hacker would want in the future.

[OgreBattle]

Thanks, I don’t know the specific terminology you use but I get the meaning

So then how does a lancer Hacker over come Mega Corp defenses, the lancer just has...

Portable cyberdeck?
Desktop computer in a possessed spirit hybrid van?

Deal with megacorps that arsenal gear sized specialized computer gargoyles and so on? Wouldn’t their hardware being bigger and more advanced mean the hacker can’t win? Like a pistol vs a battleship

[Foxwarrior]

To defeat a battleship with a pistol, you board the battleship and make a few key shots and maybe flip a few levers with your other hand. The metaphor extends to hacking too I guess, although attacks where the strength of the attacker's computer matters can easily be less dominant.

[Grek]

So then how does a lancer Hacker over come Mega Corp defenses, the lancer just has...

Human brains have better processing speed than any analog computer, by several orders of magnitude, meaning that a hacker with a direct neural interface can easily overcome the defenses of any automated system. The only meaningful defense against human hackers is a human employee who can fight back. And while the hacker can definitionally afford to attack any one vulnerability of their choosing, the corporation cannot afford to fully defend every possible vulnerability with a human employee. In practice, this means the hacker has a certain amount of time to do things before the cyberspace (and possibly meatspace) police show up to stop them.

[pragma]

So then how does a lancer Hacker over come Mega Corp defenses, the lancer just has...

Human brains have better processing speed than any analog computer, by several orders of magnitude, meaning that a hacker with a direct neural interface can easily overcome the defenses of any automated system. The only meaningful defense against human hackers is a human employee who can fight back. And while the hacker can definitionally afford to attack any one vulnerability of their choosing, the corporation cannot afford to fully defend every possible vulnerability with a human employee. In practice, this means the hacker has a certain amount of time to do things before the cyberspace (and possibly meatspace) police show up to stop them.

(Unrelated, but I'll take any excuse I can get to wax eloquent about computing in non-silicon devices.)

Brains don't have better processing speed than analog computers. Processing speed is a bit difficult to define in this context, since analog computers don't process operations in the same way digital ones do: they solve (often non-linear) differential equations and with modern technology you could make them do that quite quickly (order of nanoseconds). Brains on the other hand ARE arguably digital processors and relatively slow ones at that -- neurons spike once per ms at fastest -- but brains are massively parallel and it's tough to say how many spiking events it would take a human to solve a differential equation.

(To the brain's credit, it is wildly power efficient for its computational power, and impressively plastic. Deep Blue never ran for the Russian presidency.)

Regardless, I'm sure you have plenty of personal experience of computers doing math faster than you, and the idea that there's something magical about a human brain doing math (a) stretches my imagination and (b) isn't terribly cyberpunk. However, megacorps run big, complicated computing systems that do lots of stuff, and people who are trained to secure computing systems aren't cheap. I imagine that lancers rely on sloppy upgrades, fragmented computing resources and difficult to update hardware/software (especially if you need a new physical thing to load a new program) to find weaknesses in a corporate system. Foxwarrior's battleship analogy is a good one; you can extend it by imagining that anything smaller than a cruiser isn't manned most of the time.

As an aside: I like that this model can imply that opportunities for heists are fleeting things: "we just got word that EVILCORP is patching server software, so we have one day to figure out whatever vulnerability they're covering up until HQ manages to ship enough securtiy chips to patch it"

Finally, there's also always an option to con someone who has legitimate access: social engineering is just as old as hacking.

[Grek]

The brain is the result of an evolutionary algorithm that's been running for hundreds of thousands of generations in massive parallel. As a result, brains are extremely good in the domains that they've been evolved for, in a way that is almost entirely opaque to outside analysis. A program can number crunch its way into recognizing cat vs non-cat, or even into writing lucid poetry, but it can still be fooled by a human who is intentionally trying to be confusing*. If you need pattern matching, long term forecasting, immediate solutions to complex decision trees or the ability to operate off of incomplete information, you can't just throw a thousand X-boxes at the problem and hope for the best. You need an actual brain.

Which is fortunate, because not being able to buy perfect hacking abilities at a warehouse is a necessary conceit in cyberpunk. The basic essence of punk is that everybody wants something from you and anybody might betray you at any second to get it. If the megacorps can order everything that makes you good at hacking from a subsidiary in Lagos, that's very non-punk.

*Unless you develop it into an E grade AI. But that takes just as long as training a human for the same job does and is considerably more expensive.

[czernebog]

FWIW, if this setting really doesn't have general-purpose computers running software as a thing, I imagine that people don't really use the world "computer" the way we do. It might be that a "computer" is a "dumb smart device" that has very flexible behavior but no real smarts for reasoning about the world. "Computer" might be a generic word that people are somewhat familiar with, but which isn't as useful as vocabulary that is more driven by marketing and the social utility of the device it refers to (like a "mini-van" vs. a "hatchback," which may both be referred to as a "car"). Or it might be the case that pretty much every bit of consumer technology is technically some sort of computer, and only academics who warble on about Turing machines and models of computation care to use the term in a way that would be recognizable to a person on the street today.

A note on general purpose software: I need to second Czernobog's point that it's never going away. Even in analog chips like radios, it's common to have small amounts of embedded digital for configurability, and embedded, programmable digital signal procesing is ubiquitous.

I deliberately avoided saying anything about software, choosing instead to focus on programming, making a case for how it could still be a thing in a world without software. I don't really want to get into detailed discussion about the role that software plays in various embedded/analog computing ecosystems today because I don't think it would be a helpful discussion for AT. (I would enjoy that discussion, mind you, but I don't think it would be very helpful.)

It's important to figure out how to avoid writing in anachronisms like cinderblock-sized cell phones and making a big deal out of putting modems on things. We also want players to feel like the setting's technology is a plausible extension of technological trends of today. Note that we only have to mesh neatly with perceived trends of today, not the actual truth of technology of today: it's less important to ensure that we make truly accurate projections about future computing technology. I have had to acknowledge that AT handles cryptography and relativity with a certain amount of hand-waving and "pay no attention to the man behind the curtain." I expect that the same is true of its computing. I have listed some points above about why doing away with software-defined computer behavior is helpful to gameplay. It would be nice to flesh out more of the actual hacking minigame/cyberspace environment to see how on the mark those points really are, but I'll take them for granted for the rest of this post.

In this future setting, I've been imagining that something like Facebook is not a Web page you visit (with behavior defined in software and executed inside a browser) or an app you run (which, again, has its behavior defined in software). Instead, accessing Zuckerberg's consensual hallucination (Experienced daily by billions!™) involves slotting a Facebook chip into your deck and waiting for its drivers to negotiate with the new chip, or simply buying a single-purpose Facebook Access Device (Wondering where all your friends have gone? Catch up with the FAD™ today!) and sticking its trodes onto your forehead.

We are familiar with this model for distributing programs because that's how older-generation console hardware worked: you bought a cartridge, slotted it, and started playing your game. Anyone who has been paying attention to real-world trends in gaming knows that trends are in favor of doing as much as possible in software, to facilitate frequent patching, last-minute distribution of content, and end-to-end control of content (with DRM disallowing resale and enforcing subscription-based consumption). The AT setting will have to explain why things have shifted in favor of defining behavior in hardware. If you posit that single-purpose hardware devices are cheap to build and ubiquitous, that goes a long way. Megacorps can force consumers into paying recurring fees by selling hardware that can't be upgraded to accommodate changes in messaging protocols or by making it impossible to recharge a device.

It's very easy to have small programmable cores embedding in almost anything now, so I don't see that going away in the future. That said, having them be so slow as to be irrelevant is a good setting conceit, and their presence gives a hacker some way to modify/interact with specialized chips that would otherwise have their functions fixed at the factory.

I agree that it is very hard to get rid of software-programmable behavior. You might have small, programmable controllers or CPU-inna-CPU (like Intel's Management Engine). But even if those are locked down, secured, and laden with sci-fi DRM that actually works, you run into useful software-defined behavior really easily. If the Facebook Metaverse allows you to customize your avatar's clubhouse using what amounts to a minimal, CSS-like markup language, then it would be very cyberpunk to crack a megacorp firewall by sneaking a payload that exploits a bug in the (hardware-defined) interpreter for that language into a popular celebrity's clubhouse posters the night before a big concert. Then you can make a run at the compromised corporate ICE while hundreds of millions of adoring fans all run that markup code while they watch the Juston Blieber True Fan Concert.

If we are forced to admit that programming behavior in software won't ever really go away, we are left with the problem of explaining why software running on general-purpose processors just isn't useful in AT. This gets back to OgreBattle's question about what makes hardware-defined programs so special. One way to deal with this is to wave our hands and say that general-purpose CPUs suck, software is too slow, and you should just get on with the damn game. Maybe we can do better.

Something I don’t have a good mental picture of is how much more powerful is the specialized chip vs general hardware running software. Say sapient AI being unique hardware brains is cool, but for things beneath that.

We all agree that specialized processing units will be faster in-setting, with hardware-defined programs being faster still. What we really need is a limit on how powerful general-purpose computers are and a justification for it. Some suggestions (not all of which are compatible with AT's current "hardware slow, software fast"):

• Mind-machine interfaces are very processing-intensive. You could hook a general-purpose CPU to a set of trodes and drive an MMI through software, but, even if the software is written by a very skilled programmer, the experience is terrible. Since everyone wants to use MMI to interface with their devices, doing things in hardware is a must. This does not address the question of having software-defined control of a hardware-defined MMI stack (so your hardware acts like a GPU, providing a set of MMI interface primitives to the software). I don't have a good solution to this.
• When you're running up against corporate ICE, you can't have a general-purpose CPU slowing you down, and you need the edge that running everything in hardware gives you. The exact size of the gap between the speed of software and the speed of hardware isn't particularly important. What matters is that hardware is faster, and selective pressure has weeded out everyone who tried running up against bank ICE on a software-based cyberdeck. This is hand-wavey but very genre-appropriate.
• The war on general computation has been won by megacorps. You could have general-purpose computers running software-defined programs at modern speeds if the megacorps let you. But allowing members of the general public access to general-purpose computers is too dangerous/threatening to entrenched interests. There is a cartel that aggressively ensures that only single-purpose computers with hardware-defined behavior are manufactured in significant numbers. The specialized knowledge and industrial technology needed to construct state-of-the-art general-purpose computers are guarded very carefully. They even keep tabs on hobbyists who build 8-bit CPUs in their spare time. I'm not sure how to square this with ubiquitous 3-D printing, other than to suppose that consumer-grade 3-D printers just aren't capable of building up a processor that is more powerful than, say, a 1990s-era Pentium, and high-quality industrial or medical-grade printers are carefully policed.

You can also invent other reasons why hardware-defined programs are a necessity. Maybe, sometime in the 2040s, artificial life blossomed in software and overwhelmed the world. Ecosystems of dumb but virulent viruses, phage programs, and parasitic worm analogues overran the soft, lumbering behemoths of the old, software-based Internet. Hardware-defined programs are the solution to this that the world standardized on. Perhaps hardware-defined programs are carefully vetted so as to be inhospitable to the generation of a-life. Perhaps every program cartridge has a custom-grown CPU so that its code is mutually incompatible with that of every other cartridge. I'm not sure how to square the possibility of this kind of a-life explosion with ubiquitous computation -- you could have that explosion happen again, but at one level up from the bare metal. This would allow you to do SR-style Matrix Crash "resets" of the cyberspace/hacking sub-setting.

[OgreBattle]

The old internet being full of virus digimon and general hardware is easy for them to possess is cool imagery

[Mord]

I imagine that lancers rely on sloppy upgrades, fragmented computing resources and difficult to update hardware/software (especially if you need a new physical thing to load a new program) to find weaknesses in a corporate system.

[...]

As an aside: I like that this model can imply that opportunities for heists are fleeting things: "we just got word that EVILCORP is patching server software, so we have one day to figure out whatever vulnerability they're covering up until HQ manages to ship enough securtiy chips to patch it"

Was Capital One Hacked Or Breached?

The financial services giant says an individual was able to exploit a "configuration vulnerability" to steal sensitive records held in an Amazon Web Services (AWS) database. Experts note AWS leaks are common in cybersecurity circles as researchers routinely discover—then responsibly disclose—instances of data being left on the cloud without adequate protection.

[...]

"I wish that this case were unique. Unfortunately, these situations—where an application guarding an AWS instance is misconfigured—are all too common. In 2018, it was popular for hackers to use misconfigured 'Docker' containers in order to gain access to cloud instances. Hackers would then use those instances to mine for [the cryptocurrency] Monero."

Life imitates art... (Except for the part where art suggests that everything is done on specialized hardware in highly secure facilities, while life replies by having everything run on virtualized servers hosted on giant farms of general-purpose hardware in climate-uncontrolled barns in the cheapest backwaters that still have cable.)