I have an AI character posted <<here>> and could do with some clarification from someone who knows/understands the rules better than me...

Her home node is a humanoid avatar drone (to benefit from response increases to melee combat of all things...it's a character thing too)

When she's on the matrix I assume I use her derived AI stats? Do I run her programs on herself or does she need to run them on a commlink like a hacker? Can I run programs on her homenode (even though it's a drone?!?)...I guess autosofts work when she's not jumped in and don't when she is...?

Thanks

AIs run programs on their nodes just the same as hackers with personas do, even on drones.

To use autosofts yourself you need the Piloting Origin quality.

You use your derived attributes, limited by the node in the normal way (IIRC).

Don't forget that you can always move to another node and run your programs there instead - you are not tied to your home node like a hacker is to his 'link.

I think I follow that - essentially an AI is a bit better than the node they are in (?!?) Do the programs I want count towards that node's processor limit for response, if a commlink node I'm accessing hacking has 3 progs already running for example?

Actually perhaps I don't get it, I can use my AI stats as a node and/or another node?!?

Sorry...

Yes, everything which is not an inherent program strains the Processor Limit just like programs run by an agent.



No. An AI has System and Firewall stats because it needs them for defense/soaking, but don't they allow you to actually run programs or anything.

Not by default. They might have better inherent characteristics, but if the node in which they are running has a lower System rating then your System rating will be capped by it.


You almost never need to worry about the processor limit of a node you are accessing/hacking. You run your programs on your own node. If the target node is overburdened by IC then it reduces the response of that node, not yours.

Ok, so for simplicity I run the matrix through my commlink as a hacker would. That I can cope with. I have a firewall of 4 and the 'link has a 6...I assume for matrix combat I use the AI one and for anyone trying to hack the 'link I use that one? Seems reasonable... thanks!

Sorry, but you should really read the AI rules before causing confusion with guesstimates:
- AIs have a System rating which is independent of the node they are currently in, just like agent have. This determines the size of their condition monitor and is used for damage resistance and other cases were a system rating might be called for
- Also just like agents, AIs do not have a persona generated by the device through which they log in. They move into a node completely with all their programs and do not just project a persona icon there which uses the stats of their "home" device.

Aria, do you have access to Unwired and Runners companion? I could send you some snapshots of those pages in the books if you need them?


Sorry if that is not allowed. I purchased the legal pdf and book if that makes any difference.

@Friendbot: Got both books thanks...it's not the reading that's the problem, more the understanding of the practicalities More browsing through UW to do clearly!

@Sengir: by 'all their programs' I assume you mean inherant ergonomic progs plus whatever conventional programs they are running? Those conventional ones are being run 'by the AI' then and not on a commlink, like an agent bouncing off on an independant mission (so the AI's response is limited by the node they are in and potentially reduced by the number of programs they bring along if they exceed the AI's system -1)?

Maybe I should just stick to rigging, at least I've got those rules clear in my head

Nooooo, I want you to play Shrike darnnit!

Actually I don't thing this is quite true. In Unwired it says that the AI optimizes its home node to be better than the original stats. There is a table on Pg. 90 that shows by how much the stats of the node increase due to the AI's presence.

@Aria
An AI's inherent programs all run with the ergonomic option and do not count towards the nodes processor limit according to Unwired. Any other programs do and have to be loaded.

It says that you don't use mental attributes for Matrix tests, but doesn't say you use them for any other type of test so the logical conclusion would be that you use your Response for other tests that don't apply to the matrix. Since you will be rigging into the drone you should have no problem with that.

I am just going off of what I understand the rules to be in RC

Correct


The the programs and the AI itself run on the node, not on the AI (the AI running on itself would be weird, right? ), so the System rating of the node should be what matters for Response limits.


How does an agent control a drone, then?