Ok so by RAW it is explicitly stated that there will never, ever again be the uber scale AIs of the past like Mirage, Megaera and Deus.

Can someone please explain to me why not?

Ideally with an explanation over and beyond "because RAW says so"... ?

- J.

Under ideal circumstances, humanity would have learned from its mistakes. In all honesty, ANY A.I. given a large enough computer network to play in should eventually be able to grow to those levels on its own.

Your original AIs evolved from extremely complicated semi-autonomous knowbots (SKs) - the forerunners of what in SR4 are called agents. Many of them had dedicated home systems that were the most powerful and complex computer systems on the planet. Individually, the "Big Three" had abilities that ranged from the upper limits of SKs to almost unmeasurable - because those three were unique. There was nothing like them before, they were a new paradigm. Like when the first cyberterminal users were able to waltz over existing security like it was nothing, those first AIs had the capacity to bend or break the rules.

Comparatively, the AIs we have now are smaller, fragmented things. Their intelligence varies in awareness and comprehension, their resources are much more limited, and they have active competition now. Even if an AI managed to establish itself in the ACHE and fully utilize the technology once used to support Deus, the AI would not be Deus; it's would lack one or more of the essential traits of that massive and powerful AI, and it has limitations that Deus might not have had, in terms of imagination, empathy, and the ability to emulate or stimulate Resonance abilities.

Isn't there a line in the SR4A book that says there is hardware built into the Matrix system to stop super AI's or something?
I know that doesnt stop super AI in a private network to learn and grow though.

So long story short, its 'practically' impossible for an A.I. to grow to those levels, but not 'actually' impossible?

I would think nothing is "Actually" impossible. If your storyline needs a super powerful AI like Deus then your storyline should get a super powerful AI... You could call it Deus X (yes bad pun)

Am I now allowed you hate you, and everything you stand for?

The big three AI's where from three massive programs run on the matrix 1.0. Programs these days are tiny and specialized, made to plug in and out of handheld comlinks. If one of the Mega's had a program that big, it COULD become sentient, but i'm sure that would never happen... unless it did.




Tylonol PM makes that very funny.

The old AIs also were 'practically impossible', in all the years and despite billions sunk into research programs only three of them emerged, because AIs only formed by sheer chance from extraordinarily complex programs running on systems with vast resources.


In the new Matrix AIs can form in more usual systems and are weaker, but that does not mean that an environment similar to those which gave birth to the original AIs could not create an AI which is accordingly more powerful. However, I expect the Corporate Court to have some kind of Turing Police watching such systems, under the guise of protecting vital infrastructure from terrorism. Those folks enjoy pulling the strings of mankind from their lofty space station but damn near lost the reins when Deus decided to play god, they will certainly make sure that this will not happen again.

Artificial Resource Management is a division of the Grid Overwatch Division, under the auspices of the Corporate Court. Yes, the ARM of GOD. Yes, I am very bad.

Not Too Bad... That is awesome...

Keep the Faith

From a gaming perspective, the old AIs were plot elements first and foremost. They did not have any stats and therefore could not be interacted with on the gaming level. Whenever they appeared the players were forced to sit back and listen to whatever story the GM was telling without having any chance to interfere. This is bad RPG design, IMHO. Of course you might want to tell a story where you need an invincible super AI, but there is nothing stopping you from creating a rating 30 xenosapient AI with every program and option in the book. If it smacks the players around the matrix they will know that there is at least a theoretical chance to bring that thing down through normal gaming, instead of waiting for the narrative to offer them the "You win" choice, hoping not to die until then.
This is what I pay professional RPG authors for, because I can make up invincible plot elements myself any day.

Oh, I dunno. I don't see them quite like that..

I don't think it's necessarily bad not to assign them stats like a character. Think of them as whole castles, not individual guards. Chopping at the castle with your broadsword would be silly.

Taking on the AIs is possible, but they need to be interacted with on another level. By locating their servers and taking those down, by blocking their critical matrix backbone channels.

It's true that the old stories didn't do this extensively, but a prewritten adventure can only go in so many directions at once.
When Deus controlled the arcology, the goal could be to locate his computer core or memory banks and fry them. In the Network period, identifying and eliminating members of the Network is a possibility (which was used by NPCs).

As for the waiting for the "you win" option: some groups are more pro-active and try to develop their own ways of dealing with a hard-to-touch enemy. It requires a GM who can cope with sudden changes of plan though, and it's generally beyond the scope of a published scenario to plan for all such possible player initiatives.

But at least officially they are only policing existing AIs, not strapping explosives to places which might spawn a new Deus...or did I miss that part?


No shit

"I myself shot six toasters..."
- Lazarus Churchyard

Hm, sounds like a good title for a new .pdf release. Who wouldn't want stats for kitchen appliances?

I always thought that "Toasters" was another derogatory name for a Jarhead...

Keep the Faith

Nah, we already has stats for those.

Sorry... My Mistake...

So... what is the Challenge of wasting Toasters? Are they extremely vicious? Do they roam in Packs? And what would you call a Pack of Toasters anyway?

Keep the Faith...

That's exactly what I'm talking about. If Shadowrun would give you the option to command whole armies including trebuchets á la Exalted this wouldn't be an issue. However, the broadsword is all you'll get by game definitions.



Now you're using the reasoning of the current rules. These tactics were not feasible by the old fluff. Deus was whatever and wherever the GM wanted it to be.



As I said, if your story calls for an invincible AI you can always go for one. However, your players will never be able to go for their own plans but will always have to relay on the options your story offers to them.



And the rules as they are right now cater to this. I prefer them.

That being said, I <3 Deus and have just resurrected it in my campaign. However, I never let Deus appear as itself - its will is only felt through proxys that have stats and can be dealt with on the gaming level.

Because then we wouldn't be in Shadowrun anymore.

C'mon, you can do better than that.

Fractal Intelligence Negation Group, Extreme Response section of the Grid Overwatch Division. FINGER of GOD

*Facepalm*... *Shakes Head*...

Keep teh Faith

What about GOD's techsupport division?

Department of Infowarfare, Cybercrimes Knowledgebase.

There's nothing wrong with having a stat-less entity. One of the main themes of cyberpunk is that you can fight against the man, but you can never really win. Also, just because an entity is named, doesn't mean you have to fight it directly. "Deus" was an entity like "Aztechnology' is an entity. You can't just go directly fight against the Azzies just like you couldn't enter into matrix combat directly against Deus. The AI's were primaily story-telling elements, and that's not a problem because the AI plot arc was one of the most impressive plot elements I've ever encountered in any RPG anywhere.

Well said. But to look at it another way...



This infers almost that in order to stimulate Resonance abilities, that AIs are actually Resonance beings - rather than man made constructs like the AI of old. So in other words, there are two types of AI - manmade or Resonance.

With that in mind, surely it would be possible for man to recreate similar conditions or possibilities for another manmade AI to emerge,
Also, alternatively, a Resonance based AI could further evolve into a super AI - for example using Codivore, consuming other AIs and growing (kinda like the MCP of Tron).

The two different types obviously have different constraints in their evolution/appearance but I guess what I - and Mordinvan are trying to say - is that it isn't out of the realms of possibility.

Cheers/thanks

- J.

- J.

Jarheads? I thought Cylons, wrong universe though.

Apart from the whole "if you give it stats, the players will try to kill it" argument I thoroughly believe in theoretically killable NPCs.

As for the Siege Engines, it wouldn't be Shadowrun anymore, but those aren't the only ways to bring down a fortress. Sapping is more shadowrunlike. For example the PCs could take out the backup generators and then cut the power. No Power, no AI. Timing howver would be a problem so that the AI cannot escape to another node.

Huh? Players can only rely on the options the story offers them? I don't think my players would take that comment very seriously. Just because I'm often ahead of them in thinking of possible solutions and countersolutions, doesn't mean they don't occasionally plan in a wholly different direction.

And if course I let them; I try to make everything in the setting behave reasonably, so you can do things with it that I never expected. If they come up with a plan that can work to do something they weren't supposed to, that could work.

I don't really see how this has all that much to do with rules though.





My thoughts exactly. Taking out an AI could be precisely the job for a prime runner team; you need very special B&E and computer skills, and the level of magic runner teams boast could likely put the implant-Essence-loss-hamstrung minions of the AI to shame.

To make clear, I just see all powerful AIs as, well, deus ex machina for want of a better term (no pun intended btw) - a McGuffin or similar way to drive the story.

Hell, I was thinking just the threat of one of the old school AIs possibly emerging would be sufficient.

- J.

That's what I'm talking about: The difference between a narrative element and a gaming element. None is essentially better than the other one because they're just tools for storytelling. However, I prefer AIs to be presented as a gaming element because I can always opt for turning them into plot elements if this is the story I want to tell. This just makes you more flexible without losing anything.




Playing fair and reasonably is the mark of a good GM. However, not every GM subscribes to this (or has the skill). There are groups with a very antagonistic play style that emphasizes the gaming aspect, where it's essentially players vs. GM. Yet even if you do it differently you're still the last arbiter in all things Deus. A player might employ a tactic against the AI that worked perfectly in another campaign under another GM that you simply outrule because your Deus works differently. If your players' definition of Deus differs from yours they will always be overruled by default.

Once again, I don't want to say that this is a bad thing. It all depends on the story you're telling. An AI with stats just gives you more story options.

I think what AH was saying is "the new AIs are far more limited than the old ones, because things like imagination, empathy, and the ability to emulate or stimulate Resonance abilities are far less developed, or even missing completely". The old AIs had could interact with the Deep Resonance and even create Otaku, the new ones have no connection to the Resonance (at least no more than a mundane person is connected to astral space).

There is an obscure quote in Emergence about possible Resonance abilities among the new AIs, but that idea was probably dropped because a resonant AI would be 99% like a sprite.

I just wanted to throw this out there... current AI's as in runners companion just feel very limited to me even as characters.

The cap on attributes is just really low, limiting advancement.
Innate programs are nice... but still again, nothing a normal decker can't buy. (optomize + ergonomic runtime limits still apply)
They don't benefit from cyber/bio enhancements like a meat decker.
Then you get people who don't want to give the things like the +2 for 'hot sim' despite them being in their native perception mode.
Program costs are so rediculously low... that it's trivial to max out starting programming resources w/ basic rules (even worse if you allow people to start w/ black market craced versions).

That much said, as complimentary NPC's they seem like they could be rather usefull.
And an AI could easily have a day job quality w/o too much hassle as it doesn't need to sleep... yeah my 'day job' is being a half-time programmer for this warez group cracking/writing software on the net using my sat uplink.

If you treat the AI as just a virtual hacker, this has been our experience true.

Making full use of drone bodies however opens up some interesting possibilities.

- J.