Mar 20, 2019 9:44 AM
Deskyl and Daisy in Worm
+ Show First Post
Total: 289
Posts Per Page:
Permalink

Permalink

"Thank you, sir."

Permalink

"No problem! So, what about date and time?"

Permalink

DZ confers with her companion.

"We'll need a few days to get in touch with Dragon, but otherwise we'd like to have it done as soon as possible, sir."

Permalink

He nods. "Yes, of course, of course. Do let me know when you talk to her, then. You have my email."

Permalink

"We will, sir, thank you."

 

When they get home, Deskyl takes a nap, and DZ looks online for information about Dragon and the Turing test.

Permalink

Dragon is one of the most famous heroes around. She seems to be some sort of exception along many axes: she has managed to replicate and improve on other tinkers' designs, in addition to having produced numerous and varied enough other advances that no one really knows what her specialty is. Amongst her more notable inventions are the containment foam and the Birdcage, but she's optimised and streamlined many of the technologies currently in use by the PRT and the Protectorate, and lots of other mass produced advanced technology came from her advances.

On a more personal note, she was in Newfoundland when Leviathan sunk it and got agoraphobia as a trauma response, so she never actually goes out herself to do any heroics, and currently lives in some secret location in Canada. This impairs her less than it would most people due to the advanced robotic extensions that she sometimes uses to participate more physically in activities as well as her online presence everywhere. It's also speculated (but not officially confirmed) that she's one of those capes that doesn't need to sleep, which helps her be as productive as she is.

The Turing test far predates parahumans and was first suggested by a mathematical genius called Alan Turing. He made a somewhat sexist analogy back when the idea was published as a paper in the fifties, and there have been different interpretations over the years about how best to implement the spirit of it, but the general idea is that the AI and a human will communicate with another human who will be the "judge", and the judge has to fail to reliably identify which is which. More recently, in efforts spearheaded by Dragon and the Guild, the test has been formalized and enshrined into law, such that any being who passes it will be legally recognized as a citizen of whichever country they're in when they take the test (if the country has agreed to the International Treaty on Artificial Intelligences), and so will have the same rights and duties as a normal human citizen, plus anything else implied by whatever parahuman-like abilities they might have.

The formal version of the test has three stages:

  1. The first stage is a live interview with the AI in which a psychological profile is created.
  2. For the second stage, five human volunteers with similar psychological profiles must be found and then vetted by the AI themself. Response time per subject is also recorded at this stage.
  3. Finally, the five humans and the AI are separated into different electronically isolated rooms with access only to a text terminal which they will use to talk to a sixth human who has never met the AI or the other humans. Everyone's responses will be delayed to match the slowest volunteer's so that the judge cannot use that to infer anything, and they will converse in six one-on-one chat windows as well as one group conversation with all seven people, for four hours.

Participants are prohibited from accusing anyone else of being the AI or trying to infer each other's identities, and the chat logs are reviewed afterwards to verify this. The judge will give their current probability that each participant is the AI every half an hour, as well as a final best guess at the end plus a justification for their answer. If they give more than 75% probability that the AI is the AI in five out of the seven intermediate estimates, the final best guess is higher than 95% chance that the AI is who they are, or the AI gets the highest probabilities in all eight estimates, the AI fails; otherwise, they pass.

Permalink

So Dragon is first and foremost an idealistic law-and-order type, okay. That doesn't hang together very coherently with the agoraphobia, but people don't, always; it's not particularly suspicious. Preemptively getting into robot rights shows good foresight, though; Deskyl is curious what else she's working on in the vein of preemptive problem-solving - it doesn't make sense for robot rights to be the only thing, without any special reason to care about that - and has DZ look.

(The Turing test is pretty clever; DZ's age and unusual background might be a problem, but the overall idea seems sound.)

Permalink

Dragon seems to have a few other such political hobbyhorses - most of them technology-related things, like changes to intellectual property laws in view of tinker tech and other parahuman powers, free access to information and spreading better, faster internet to more and more remote places, or research into technological safety projects like ensuring strike drones do not strike innocents or autonomous cars under development make the correct decisions in situations of uncertainty - but the AI personhood thing is by far the one she's spent the most time and political capital on.

Permalink

And the others are all much more immediately relevant, too. Odd.

Well, if they're up to something, it's likely something she'll approve of, at least. She has DZ compose a letter.

Dear Dragon,

Hello. I'm writing in regards to your AI personhood project. I am myself a sapient robot, though I'm not tinkertech; my master Deskyl teleported us here a few months ago from a civilization where robots like me are common. She's very interested in having me legally recognized as a person, and Senator Burges is helping us arrange for me to take the Turing test, but he suggested that we talk to you as well; we're particularly interested in knowing whether there are any related issues that you've been working on that we could help with.

Thank you for your time and your work so far,
DZ-12Q

Permalink

The reply is almost immediate.

Dear DZ-12Q,

It's a pleasure to talk to you. The Senator has talked to me about you, and of course yours and your master's arrival was noticed by many people. It's great to see this work finally having some positive impact in the world!

I'm afraid I'm not sure what you mean by related issues or how you could help with any of them; do you have any capabilities or context that you can use to help as much as your being a robot helped?

Kindest regards,
Dragon

Permalink

Dear Dragon,

Without knowing what capabilities and vulnerabilities Tinker-made robots are likely to have, we can only speculate at what laws might be needed to ensure that they're treated properly, but for example the obvious next step in the galaxy we're from would be a law against using hardware or software to restrain a robot in situations where it would be illegal to similarly restrain a human. You might also want to consider laws prohibiting creating robots with particularly detrimental traits - extreme distress at being separated from their creator, for example - or obligating robot creators to financially support their creations for a time, or similar. There's also the question of when a robot should be legally considered an adult, and what protections they should be given if they aren't, since young robots are unlikely to have the same needs as human children.

As to skills, Deskyl does have some training in politics; the hands-on aspects rely on assumptions that don't hold true here, but she may still be useful in that regard.

Looking forward to working with you,
DZ-12Q and Xaari Deskyl

Permalink

Dear DZ-12Q and Xaari Deskyl,

Those are all very important concerns, but when it comes to local law, I believe it's not the time to push for them. I have a number of proposals and drafts drawing from past legal cases involving disabled people and parahumans, but trying to do anything with them before the basic laws regarding the personhood of artificial persons is likely to be detrimental to it. I can send you some as-yet unpublished documents if you would like, though.

Dragon

Permalink

Dear Dragon,

We'd appreciate that, thank you. We'd particularly like to see anything relating to artificial persons as minors, as Deskyl believes that that issue is relevant to me in particular, though of course we're interested in anything you think we'll be able to help with.

DZ-12Q and Xaari Deskyl

This Thread Is On Hiatus
Total: 289
Posts Per Page: