Comic Discussion > QUESTIONABLE CONTENT

WCDT Strips 3631 to 3635 (11th to 15th December 2017

<< < (18/58) > >>

dutchrvl:

--- Quote from: BenRG on 12 Dec 2017, 04:00 ---
It's an interesting thing because, up until very recently, Winslow was the controlling intelligence of Hannelore's PDA. So, it is very possible that he has read every email sent to her and that she has sent for years. It may literally not occur to him that there is a limit to the access he should have to her personal communications.


--- End quote ---

This brings up an interesting question: when AIs first function more-or-less as their companion's PC/ipad/smartphone, do they actually have access to email/browser programs etc., or do those function as standalone software that is protected from access by the AIs themselves (e.g. the AIs can help opening the email program, but not access the content)?

Especially once an AI companion decides/gets the opportunity to transition to a humanoid chassis, its role is bound to change considerably, for instance (even more) towards that of a typical human friend. With that change in role, the high-level access to much of their human companion's private matters during their previous incarnation could lead to problems down the line.

Think about it: you make a new friend, but for the past years they may have had access to all your emails, web browser history, photos, music, etc....Would that make you comfortable?

I seem to recall an earlier comic in which pintsize mentioned that Marten does not use him for web browsing for exactly the above issues, but I haven't found the comic yet.

tustin2121:

--- Quote from: BenRG on 12 Dec 2017, 02:22 ---I do have to say that today's strip does make me wonder that, long-term, Tilly's role will be as Winslow's significant other rather than anything directly to do with Hannelore.

--- End quote ---

Please no. I actually like Winslow, and I want it to stay that way...

Bad Superman:
Regarding the AI discussion:

Some of you might be familiar with the GURPS roleplaying system.
One of the sourcebooks of this game is called "Transhuman Space". It's an actually very well crafted soft sci-fi setting, meaning, it's not so much focused on all the cool new technology (that would be the playing field of hard sci-fi), but more on the way societies change due to vast technological advances and breakthroughs. (The meaning of 'sex' and 'gender' for example changed quite a bit, in a world where true "Morphological Freedom" is possible. With enough money, you can have exactly the body you want. There are no limits.)
AIs are one such breakthrough.
In this fictional background, AIs can have various legal statuses, reaching from 'free citizen' to 'second class citizen' to 'valued property' to 'abomination'.

One other drawback of being an AI in this world is their "Reprogrammable Duty".
It's an optional disadvantage for AI characters in the game. They're still software after all. If, as an AI, someone has your command codes, you obey them, no matter what degree of freedom you might otherwise possess. And even if you are aware of these codes, there are safety measures in place that prevent you from changing them, or acting against the people who know them. Only 'free citizen AIs' have full custody over their own command codes.


I wonder how the AI characters in QC are programmed in this regard. Especially AIs with military-grade bodies like Bubbles. From what we have seen until now, AIs in QC are truly free. Free enough at least, to create their own illegal robot fight club, like Corpse Witch, or download tons of porn like Pintsize...
Could AIs like Winslow and Momo still have some limitations written into them, that keep them from invading their human companions private data? This would add an interesting new dimension to Winslow contemplating the fact, that he is now physically able to open and read Hannelore's mail if he so chooses.

Is it cold in here?:
Jeph once said something like "absolute free will".

I hope Winslow does not go to Robot Jail for mail tampering.

Thrudd:
That is an interesting hypotheses.

Taking what we sort of know about AI conception I very much doubt they have any third party additions to their being per se. Control codes et al.
Station, Spooky and the rest of the advanced AI, as well as a number of influential humans would not look too kindly on such chains and shackles on any recognized intelligence.

Now as to limits on behavior and such, when in specific chassis or functions, that I can see having built in safeguards.
When you think about it, it makes sense, just like OSHA and WHIMIS are out there to safeguard operators and those in specific work areas from hazards intrinsic to a job function.

In other words, lockout and key codes and such would be built into the chassis and not into the "person".  (click to show/hide) let us not get into schlock mercenaries memory rewrites, effetive imortality and such for now - that is a whole different ocean of fish - ocean because a kettle just aint big enough Like the ignition key for a vehicle - anyone could hop in and look around and fiddle with the dials and buttons but the one with the key is the one able to start the engine.
The other end is military hardware where there is a command from higher up authorizing / unlocking weapon systems and ordinance.
The ground pounder may be able to use their sidearm or punch someone upside the head at their discretion [answering to the brass after the fact] but needs authorization from HQ to put that SAM to good use.

Navigation

[0] Message Index

[#] Next page

[*] Previous page