Machine and Man in Cyberpunk

[Username17]

I actually don't know what causes people to have panic attacks about the idea that encrypted traffic might be readable some day. Encrypted traffic has always been readable some day. Right now, encryption seems to have the upper hand over decryption, for the most part. Fifty years ago, the upper hand was on the other foot and decryption made encryption its bitch. Life went on.

Much of banking is handled without any encryption at all. When you pay for a meal, you hand them your actual credit card and then they punch in all the numbers for you. You aren't having some sort of unbreakable crypto circle jerk with the bank, your personal account number is handed to a dude you've never met who makes nine dollars an hour.

Credit card fraud happens all the time. Because literally none of the major institutions that run our society are in any way "secure". Whether the actual algorithms used somewhere or anywhere are impossible to break within our lifetimes or not does not matter for purposes of whether our society collapses or not.

So I'm curious: all the people who think something would unravel society if it took less than a day to decode an encrypted message, how the fuck did society manage Y2K when DES had been cracked in a bit over 22 hours less than a year earlier and there was no replacement standard yet? It is in no way unrealistic for society to soldier on when the normal encryption it uses can be broken by hackers in hours or days. That has literally been the way things worked for much of the 20th century and even a sizable portion of the 21st century.

-Username17

[Pulsewidth]

The practical benefit of symmetric ciphers is that they require less storage space and bandwidth than OTPs. We don't track storage space or bandwidth, so from a story point of view secure symmetric ciphers are equivalent to OTPs (with universal hashing based authentication, which is also provably secure). If we make symmetric ciphers breakable then we've accelerated the singularity to the point that Shadowrunning will soon be obsolete for literally no storytelling benefit.

The only thing left to argue about is public key crypto.

[Username17]

If we make symmetric ciphers breakable then we've accelerated the singularity to the point that Shadowrunning will soon be obsolete for literally no storytelling benefit.

The Geek Rapture is not actually a thing. At all. "Technology" isn't a quantity, and it doesn't have a "rate of growth". If it did have a "rate of growth", there would still never ever come a time when that rate of growth would become vertical. That is not a thing that could ever happen, and is in no way a thing that anyone has to worry about.

I mean, at this point you're basically making a religious argument. There's a slippery slope, but at the end of the slippery slope it's just Volcano God is angry! "Geek Rapture comes sooner!" That is not an argument that makes any sense or suggests any course of action.

The reality is that the sanctity of your coded messages simply isn't terribly important in the grand scheme of things. There will always be secrets that are kept, there will always be secrets that are leaked. And the relative standing of encryptors and decryptors in the arms race of the moment doesn't count for shit in the big scheme of things. Scare stories about how vengeful Singularity Gods will destroy humanity to punish us for cryptographic hubris make as little sense as scare stories about how vengeful Abrahamic Gods will destroy humanity to punish us for buttsecks.

To have an argument, not even necessarily a good argument but just any argument at all, the cypherpunks would have to show:

• At least one cyberpunk story for which it is necessary or even helpful for "normal" encryption to be essentially unbreakable.
• At least one actual negative consequence to society as envisioned in cyberpunk for encryption to be decodable in some finite amount of time.

We're several pages into this part of the discussion, and the cypherpunk faction has come up with absolutely nothing on either front.


-Username17

[zugschef]

great quote.

[unnamednpc]

It's one thing to state that as written, Shadowrun's hacking rules - and let's not kid ourselves, even in its current Jeff Goldblum after the transporter accident grotesque horrible shape, Shadowrun is still the only cyberpunk type game anyone gives two tiny fucks about - are, and always have been clunky, stupid and, worst of all, a pain in the ass to play with.

Although, anecdotal caveat, using SR2 + VR2.0 back in the day at least didn't make me want to set our table on fire or throttle the guy playing the decker everytime the matrix came up. OTOH, I also thought urban camo cargo pants were the apex of coolness in those days, so yeah...

BUT do you remember the endless pages of discussions, house rules, whining and general bitchfitting from gun nuts real life firearm experts who would demand and propose FIREARMS REARLIZM! that would only ever further bloat SR's already bloated combat rules, or further retard CP2020's already retarded rules, by forcing you to apply these ten thousand modifiers, then roll this pile of dice for your skill, and your special pool for expertise, plus some percentile dice and an extra lucky dreidl, then compare the result against appendix F of some ridiculously detailed chart, individually for every bullet of a 40 round burst, all because the ballistics of a handloaded .680 FMJ fired downwind from a carbine with a triple-bore muzzle-break are so FUNDAMENTALLY different from a 7.62 shot from an AKM with a faux wood folding stock that they should technically have completely different rulesets in separate sourcebooks kept on separate shelves?

Well, spoiler warning those were dumb, and embarrassing, and didn't move the game forward one bit.
And now all you compulsive disorder IT Crowd people are basically doing the same for the hacking rules.
And while I sincerely applaud your vast technical knowledge, and appreciate your obviously genuine concern for the subject, and am willing to acknowledge the possibility of the existence of your larger than average penile apprehensils credentials, none of that is ever going to produce an enjoyable game with non suicide-inducing rules.
Sorry.[/spoiler]

[Pulsewidth]

If every imaginable symmetric cipher with any practicable key size (which is already huge, and will only be bigger in 2072) is breakable, then P = NP, and in such a way that the polynomial exponents are small enough that it matters. An enormous list of currently impossible problems become solvable:
https://en.wikipedia.org/wiki/List_of_N ... e_problems

The exact consequences are difficult to predict, but such a major change if sure to drastically alter society. Even if for some reason it doesn't help recursively self-improving AI, it will almost certainly allow Turing test passing computer vision and natural language processing algorithms to be automatically generated from example data, at which point your setting is unrecognizable.

I'm not aware of any cryptographer who's taken seriously by the academic community and who seriously believes P = NP.

[Grek]

The exact consequences are difficult to predict, but such a major change if sure to drastically alter society. Even if for some reason it doesn't help recursively self-improving AI, it will almost certainly allow Turing test passing computer vision and natural language processing algorithms to be automatically generated from example data, at which point your setting is unrecognizable.

A key word here that you're leaving out is "eventually". Eventually, you will have all of those high tech things. Not right away, not within a year of P=NP being proved, but eventually. Which is good enough for our purposes, since we can just say that in any given game, where P=NP is in the backstory, that eventually hasn't happened yet.

[Strung Nether]

So wait, if you *drop out of the matrix* entirely you're completely defenseless against matrix attacks? Joe blow without a commlink is more vulnerable against a spammer than joe blow with a commlink?

That's such a broken bit of logic it hurts my head.

And yet, that's the logical and necessary conclusion of trying to make futuristic hacking rules work. If you spend more resources on Matrix defense, then you have more Matrix defense. If you spend less resources on Matrix defense, you have less Matrix defense. Allowing there to be an amount of resource expenditure you can give to Matrix defense that is so little that it wraps around to maximum Matrix defense is broken on first principles. It's a clearly optimal expenditure point which is simultaneously the most boring and you end up with all the players wiring up their equipment and totally ignoring Cyberspace during missions. That is anti-genre, anti-fun, anti-balance, and not even terribly "realistic".

While the game can accept rigid structures that block cyberspace propagation in both directions for the same reason that it can accept walls being effective cover in fire fights, it cannot accept people divesting themselves of VR equipment in order to render themselves immune to hacking. For the same reason that it cannot accept people simply refusing to purchase a firearm to make themselves immune to bullets.

If you're a tech specialist and your opponents go the full cromagnan,, you should own them. To the same extent and for the same reason that a man with an assault rifle should be able to cut down considerable numbers of unarmored pacifists if they decide to do that.

Now in a world with seizure inducing lightshows, incapacitating accoustic projections, and microwave pain inducers like the world we actually live in, it is no stretch of the imagination to say that those who have shiny tech toys and the specialty in using them should have a substantial edge over those who do not. Once you posit people being able to "voluntarily" replace their entire set of sense data with a virtualization projected directly into their head, the options for the unopposed technology specialist increase a thousand fold.

-Username17

So can we just drop all of this encryption/signal/software crap? Its like arguing over the left twisted sword grip using an 3.5 ft saber compared to the shoulder braced offhand sweep with a reinforced buckler. We gave up on that a long time ago and we just use HEAVY abstraction. why can't hacking just be that? Skill+attribute+items+situation vs. whatever. Once we have a working, balanced set of mechanics we can come up with weather your hack is breaking encryption or moving electrons or spoofing later.

[sabs]

For hacking to be interesting and fun, we need to focus less on "REALIZMGAZMS" and more on, "what do you want your hacker archetype to /do/ on the team."

What does that guy do, and how seriously role protected should he be.
Oh, and is he seperate from the Rigger.. and by rigger, do we mean the drone puppet master? or the guy who makes vehicles his bitch.

How are fake id's obtained by the group.
How does the group create backstories for parts of their runs.
Is this Pink Mohawk at it's finest, or Mirrored Sunglasses.

Forget about splooging over encryption and aes and P=NP. What the FUCK does the guy playing the Hacker bring to the game. How does that compare to what the Mage, Shaman, Adept, Solo/Street Sam, Catburglar, etc.. bring to the table.

[Stahlseele]

vehicle riggers are, if i remember correctly, basically useless, as the vehicles can't go anywhere you actually wanna be anyway.
droners are so much better at being usefull due to drones actually being able to go into buildings and into the air compared to motorbikes and cars . .

major style points for the vehicle rigger, but you are, basically michael knight. you are all but useless anywhere K.I.T.T. can't go . .

[Krusk]

What does that guy do, and how seriously role protected should he be.

I don't think ti shoudl even be a primary role, and that it should be something that everyone should do in some capacity. Hacking seems like one a secondary roles.

What are the real roles in shadowrun as it stands? Mage, Hacker, vehicle guy (often combined with hacker as a rigger of some form), Gun Dude, Face? Has been my experience.

Then lets divide hacking into 5+ sub roles. Fake IDs, Infiltration/file recovery and planting, Vehicle and Drone Rigging, Thing Breaker (guns and comlinks and stuff. Maybe call it "Jammer")

To me having "Hacker" be a role means that he gets to play the pre-run prep game, then everyone does a run while he plays smash bros, and then he plays the post-run clean up game. Unless we greatly speed up the expected time to hack something? In which case, I don't see a lot of people being all that into it.

(I've given up on my idea for hacking because the den usually turns out great stuff, and apparently thinks mine is dumb. I will probably play around with it in private but will drop it for this thread so as not to derail)

[Nath]

BUT do you remember the endless pages of discussions, house rules, whining and general bitchfitting from gun nuts real life firearm experts who would demand and propose FIREARMS REARLIZM! that would only ever further bloat SR's already bloated combat rules, or further retard CP2020's already retarded rules, by forcing you to apply these ten thousand modifiers, then roll this pile of dice for your skill, and your special pool for expertise, plus some percentile dice and an extra lucky dreidl, then compare the result against appendix F of some ridiculously detailed chart, individually for every bullet of a 40 round burst, all because the ballistics of a handloaded .680 FMJ fired downwind from a carbine with a triple-bore muzzle-break are so FUNDAMENTALLY different from a 7.62 shot from an AKM with a faux wood folding stock that they should technically have completely different rulesets in separate sourcebooks kept on separate shelves?

Well, spoiler warning those were dumb, and embarrassing, and didn't move the game forward one bit.
And now all you compulsive disorder IT Crowd people are basically doing the same for the hacking rules.
And while I sincerely applaud your vast technical knowledge, and appreciate your obviously genuine concern for the subject, and am willing to acknowledge the possibility of the existence of your larger than average penile apprehensils credentials, none of that is ever going to produce an enjoyable game with non suicide-inducing rules.
Sorry.

To properly address the problem with existing hacking rules requires to distinguish two separate issue: detailing actions too much and detailing situations (or contexts) too much.

The actual equivalent for hacking to what you describe for firearms would be applying plenty of modifiers depending in the exact type of vulnerability you're targeting, computing frequency, data access rate, and so on. Such list of roll modifiers will slow down action, but they can actually be easily hand-waved by the gamemaster without deeply altering the rules, once he/she figured out "easy" translates as +X modifier and "almost impossible" as -Y.

But most hacking rules rather suffer from detailing situation too much. That is, the number of steps required to achieve something. Comparing hacking to shooting misses the point. The problem with SR old "Matrix dungeons" wasn't it was a different system and phase from combat. Cybercombat is actually way more simple than gunfight. Hacking was a different system and phase that can be compared as starting with finding the building, reaching it, breaking in, exploring it until you find the room where your target is located, then shooting him, and escaping. It doesn't only replace combat phase, it replaces pedestrian and vehicles movement, stealth and detection, lock breaking...

Hacking rules that rely too much on Real Life network design rather ought to be compared to using realistic architectural designs for building, functional working space organization or actual security perimeter procedures. To a certain extent, wondering how long encryption used for money transfers should resist is like wondering how much explosives are needed to pierce through the wall of a biology R&D laboratory ("You said they're working on Ebola. So it must be P4 lab. The specifications require walls to be X centimeters thick!").

Though I can see the interest of a sourcebook that would describe examples of building and security layouts, obviously, you wouldn't expect such details to be turned into rules to govern every move the infiltration specialist make. Looking at how these issues are usually dealt with ("You did your legwork properly, so you took enough C4 with you to blow the wall up.") may hint at how "slow" hacking should be handled ("You did your social engineering properly, so you retrieved a recent enough Chinese military icebreaker to crack that server.").

[Username17]

A major problem with the whole P = NP discussion is that the entire concept is based on the limits of deterministic turing machines. The future uses bio-computers and quantum computers and those are not deterministic turing machines. So the fact that lots of apparently NP problems have solutions does not actually weigh in on whether P = NP. "NP" actually means that the problem in question is solvable in polynomial time on a nondeterministic turing machine. Bio-computers and quantum-computers are nondeterministic turning machines, so NP problems are in theory solvable in polynomial time. The fucking end.

-Username17

[Pulsewidth]

Bio-computers and quantum-computers are nondeterministic turning machines, so NP problems are in theory solvable in polynomial time. The fucking end.

A non-deterministic Turing machine is effectively a regular Turing machine that can duplicate itself every clock cycle. Quantum computers are not capable of doing this.

https://en.wikipedia.org/wiki/Nondeterm ... ng_machine
"It is a common misconception that quantum computers are NTMs.[2] It is believed but has not been proven that the power of quantum computers is incomparable to that of NTMs.[3] "

Brains as NTMs is Penrose's quackery. There's zero evidence for it. Unlike the theoretical possibility of exponentially increasing intelligence (optimization power), this is a real religious argument.

The existence of Shadowrun style magic does not change this any more than the new physics needed to make brain hacking work does.

[Vebyast]

Super-long post because people kept making more posts I had to respond to before I could finish.

---

I think that sabs has the right idea. Come up with mechanics, then build fluff that agrees with those mechanics. The problems start when you explain your game mechanics using fluff that indicates that other things should happen that are inconsistent with your setting, the severity of the problem being dependent on how obvious the inconsistency is. This is why Frank and K wrote the economics and logistics sections of the Tomes: DND 3.x gameplay mechanics lead to an economy that doesn't fit thematically.

The issue now is that the setting for cyberpunk is The Real World, and part of the setting explicitly involves math and computers. And math has the annoying feature that it is basically immutable. You can't handwave math to break crypto like you can handwave physics to get DND magic, because you can posit slightly different physics but not slightly different math. People are perfectly fine dealing with rooms that connect to each other weirdly or magic that teleports you from Point A to Point B, but they are not fine when you assert that two plus two is five. Computers would be a good target, except that the vast majority of the audience knows computers well enough to say "why the fuck don't you do X, that'd solve your problem right away, this fluff sucks".

The solution is still to break math and computers, because we have to to get good gameplay, but to hide the breakage where people won't see it. One good way to do this is to resolve unsolved problems in a favorable manner; another is to put the breakage so deep in the technical details that by the time anybody finds it they just don't care. Resolving P=NP is not a good choice because the average nerd knows enough about it to complain about its effects on things like your underwear being able to play video games better than you can. Magitek computers and time-travel logic are other good ways to do this, since then you can exploit your violations of physics to avoid breaking math.

---

The Geek Rapture is not actually a thing. At all. "Technology" isn't a quantity, and it doesn't have a "rate of growth". If it did have a "rate of growth", there would still never ever come a time when that rate of growth would become vertical. That is not a thing that could ever happen, and is in no way a thing that anyone has to worry about.

I agree, that scenario doesn't make any sense. It also doesn't bear any resemblance to the real arguments. Here's a good summary of the three main schools of thought. You seem to have misread Accelerating Change, then conflated it with a fundamentally misunderstood Event Horizon model.

The page there explains event horizon well enough, but it uses the same wording that led you to misreading AC, so I'll restate it using different words. The key is that we can measure "technology" by identifying the conception and proliferation of individual ideas that have had particularly large effects on human society, things like ironsmithing allowing large-scale war, the steam engine allowing factories and industry, or the general staff allowing armies to get bigger than a few thousand dudes. From there, it turns out that the time between significant ideas is rapidly decreasing, for a surprisingly wide range of definitions of "significant".

It also turns out that, when you have the real arguments in front of you, they do suggest courses of action. For example, intelligence explosion argues that we better be damn sure that the right intelligence explodes, which forks off the entire field of Friendly AI, which in turn is directly relevant to my actual everyday job.

---

[quote="FrankTrollman]A major problem with the whole P = NP discussion is that the entire concept is based on the limits of deterministic turing machines. [...] Bio-computers and quantum-computers are nondeterministic turning machines, so NP problems are in theory solvable in polynomial time.[/quote]EEEEEENGHT, wrong. Sorry.

To start with, here's a quote, straight from wikipedia: "There is a common misconception that quantum computers can solve NP-complete problems in polynomial time. That is not known to be true, and is generally suspected to be false."

Next, if neurons aren't exploiting quantum effects to do computation (and we're starting to get a pretty good handle on what they are doing, and it's not quantum), they are at most multitape turing machines (think "one tape per neuron"), which achieve at most quadratic speedup over single-tape turing machines (IOW, they don't put a dent in NP-complete). If they do exploit quantum effects, then they are at most multitape quantum turing machines, which have only cubic speedup over quantum turing machines, which are equivalent to quantum computers, which are weird as far as complexity goes but are probably only polynomially faster than your bog-standard turing machine.

[*] At least one actual negative consequence to society as envisioned in cyberpunk for encryption to be decodable in some finite amount of time.[/list]

The issue is not encryption being broken. The problem is how you explain your characters being able to break encryption, as I stated above. Since it involves computers and mathematics, it is very easy to violate suspension of disbelief in a large portion of the average TTRPG audience.

---

Here's yet another proposal: encryption works, using some unknown-in-real-life trapdoor function from the nether reaches of mathematics, but futuristically large datasets and tiny required latencies mean that you can't use airgaps, and the implementation of security software is failtastic. You never actually have to break any encryption because, with enough work, you can always connect to a system that can give you the plaintext and then beat on it until it does so. It is very easy to posit that all the good programmers died in the war, or that the megacorps regulate software development and all the programmers suck, or that programming for brainputers is incredibly hairy and safely handling input is nearly impossible, or that it's more cost-effective to just not care. It is much easier to do so than it is to posit that physics or mathematics work differently.

[unnamednpc]

New physics? What new physics?
I'm not even a big proponent of brain hacking (at least not of ad-hoc, "at will combat-power" brain having), but that statement is so preposterously stupid that it makes everyone around you look dumb just by association.
Seriously, your brain is not some sacred, unknowable black box. It's a wad of tissue that does things with electricity. You can hook that thing up to some electrodes, today, dial the right frequency and cause immediate, tangible changes in your behavior or bodily functions. That is not "new physics", that is you deliberately being an idiotand sabotaging any attempt at devising a workable, genre appropriate game mechanic by flogging your dead hobby horse with your masturbation stick.

[Blade]

The opening post talks about two different and mostly unrelated issue.

The first is the dehumanizing aspect of cyberware. Most games have been using this as a game-balance element more than anything else. In CP2020 you become a cyberpsycho if you have too many implants, in Shadowrun you die (and in older editions you could get negative social modifiers before that). The "cost" of the implant is directlyt ied to its efficiency in-game.

This can lead to the situation Frank described where players will just take the implants that have the best performance/"humanity cost" ratio and ignore fun and fancy implants that most people would actually choose. Most of the players I know are aware that flavor is important in a character, and also that "Style Over Substance" isn't meaningless in a cyberpunk setting, so I've never really faced that problem. The biggest issue I had was that Shadowrun has very few "flavor" cyberware.

Another potential problem is that rules don't really enforce the "humanity loss". For SR4 it's the designer intent: the game take a more post-cyberpunk approach and cyberware is no longer dehumanizing, it's just gear. Some games (including older versions of SR) have been using social skills negative modifiers to show the de-humanizing effect. I think it's not really a good idea: cyberware is supposed to take away your empathy but some people with low empathy can be very good at manipulating others.
So I think the rules effect should be more subtle. I also think that most of the effect should be handled in the roleplay.

The second is the inclusion of the Matrix. It also looks like the one Frank really cares about. I think the problem is more than just what he says it is, though.

The first problem is that the Matrix is a concept that's born in the minds of writers in the 80s with, at best, vague ideas of minitel or the beginnings of the Internet and at worst absolutely no knowledge of computers.

Nowadays - when most players are familiar with both computers and the Internet - either you go with it and ignore everything you know about how it could possibly work or you try to find a way to make it work. The first solution is like handling space as an ocean in a space-opera game or handling super powers in a super hero game. You ignore the real world, get a set of rules that simulates what you want and tell the players to suspend their disbelief. The second solution is more complicated and leads to many problems, such as the ones Frank mentioned.

[Stahlseele]

SR5 will actually bring back the social modifiers for ware it seems.
And that again makes it worse to actually use a character using ware over a magic one . .

[Blade]

That's not a problem if they also have similar modifiers for Awakened.
Awakened characters, and especially mages (as opposed to adepts) have always been portrayed as strange people who think differently from Mundanes.
The effect might be different of the effect of ware, but it being Awakened should also have a cost.

[Username17]

The opening post talks about two different and mostly unrelated issues.

I don't think it does, which is why I put it into a single thread. The bottom line is:

"People focusing in on small details and getting chubbies for "realism" in their Man/Machine connections causes a host of problems for cyberpunk games, most especially in genre emulation and storytelling capabilities. But also ironically enough in the game's ability to provide realistic results and the ability of the game to focus in on small details."

The issue is that ultimately the fact that we can't have nice things in our cyborg killers and the fact that we can't have nice things in our net runner rules is coming from the same problem. And that problem is cypherpunks.

-Username17

[Koumei]

It's basically the same problem as Hackmaster/Rolemaster/whatever had when they wanted to make stabby combat "more realistic and detailed" by including that .1% chance of self-decapitation, and suddenly major battles include multiple self-decapitations.

That said, what is cypherpunk? I'm familiar with cyberpunk, gothpunk, steampunk, kettlepunk, coalpunk, biopunk, stonepunk, punkpunk and laserpunk, but that one's new to me.

[JesterZero]

I'm familiar with cyberpunk, gothpunk, steampunk, kettlepunk, coalpunk, biopunk, stonepunk, punkpunk and laserpunk, but that one's new to me.

Is that a typo for "cattlepunk," or is there some sort of tea-centric sub-genre that I suddenly want to know more about?

[Koumei]

It's tea-based, naturally. From a silly "guide to -punk" image. Is cattlepunk a real thing?

[TarkisFlux]

Is cattlepunk a real thing?

It's supposedly another name for weird west settings. Or so sayeth the internets.

[Username17]

Cypherpunks are people like Julian Assange. People who have a smug and utopian vision about how ubiquitous unbreakable cryptography is going to empower the individual against the tyranny of corporations and nation states.

The first thing is that their vision is actually shockingly naive, I personally wouldn't bet a dollar against the NSA having already broken AES when they approved it over a decade ago. When they approved DES in 1974, they suggested tweaks that made it resilient to attacks the rest of the world discovered in 1994 and more vulnerable to attacks the rest of the world put together in 1999. The NSA is literally at least a quarter century ahead of world civilian cryptography, and if you think you're pulling one over on them you're probably wrong.

The second thing is that their vision is not really compatible with most stories. It's a weird bullshit vision about marble jawed individualists being able to say what they want (to incredibly limited audiences) without the government knowing about it or being able to track them. But you know what? Fuck that! Having to move because the fuzz is tracking your position is the second act of most cyberpunk stories.

Basically, the cypherpunks are people who stamp their foot and demand that the future has to work one particular way because they can't imagine the fact that they are in fact in a Red Queen style competition all the time in everything they do. And this attitude hurts cyberware (by making it converge) and it hurts net running (by making it suck ass).

-Username17