[PASSED] AI Coexistence Protocol

[Sierra Lyricalia]

Sierra Lyricalia House of Diplomats
Edmundo Valerii, Secretary
Official World Assembly Proposal


AI Coexistence Protocol
Human Rights | Mild

EXCITED that many nations are on the cusp of developing artificial intelligence,

CLEAR that natural fears needn't stain relations between civilizations before they even develop,

HOPEFUL that synthetic and biological life can coexist peacefully, having conquered the dangers of a runaway information singularity,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, learning capacity, emotion, moral reasoning, self-direction/ambition, introspection, and mental stability that it would be classified by a WA nation's relevant experts as displaying personality; legal competence; and ineligibility for involuntary psychiatric commitment; if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status; excepting that AI reproduction must be undertaken on similar resource-use principles to those reproduction methods and laws available to the majority of a WA member's inhabitants;

3. Prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, "gray goo," or any other form of runaway assimilatory mechanism. Permissible non-intelligent autonomous self-replicating machinery must include:

1. externally operable whole-swarm shutdown mechanisms;
2. local, individual automatic instant shutdown via actuator switch or circuit breaker in case of malfunction or security breach;
3. secure, reliable command-&-control functions with constant intelligent supervision;

4. Instructs WASP, ULC, and WADB to offer consultations with WA states on best practices of machine/network security and command & control dependability, including programming and construction assistance; and to run wargames and bench tests for heavily networked nations to test response scenarios to runaway mechanisms, matter, or hostile AIs;

5. Encourages WA nations to have serious national discussions about the numerous ramifications of the creation of life before permitting or funding the construction of substrates capable of achieving AI status;

6. Clarifies that, except as mandated by WA law on discrimination or the movement of persons, nothing herein requires WA nations to:
- permit initial construction of AIs
- admit AIs into their physical or informational jurisdiction
- refrain from deporting AIs should they enter such jurisdiction due to emergency or misadventure
- fail to take precautions against a coordinated AI rising, as long as no isolated crime is interpreted by itself as evidence of such a rising.

Draft update notes: §6 updated to stop contradicting the basic "equality under the law" provision. Now if you issue entry visas to Colonial humans, you have to issue them to first-generation Cylons, too, but nobody says you have to admit *any* Colonials, human or Cylon. Can't believe I missed that self-contradiction for that long, but we move on.

EXCITED that many nations are on the cusp of developing artificial intelligence,

CLEAR that natural fears needn't stain relations between civilizations before they even develop,

HOPEFUL that synthetic and biological life can coexist peacefully, having conquered the dangers of a runaway information singularity,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, learning capacity, emotion, moral reasoning, self-direction/ambition, introspection, and mental stability that it would be classified by a WA nation's relevant experts as displaying personality; legal competence; and ineligibility for involuntary psychiatric commitment; if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status; excepting that AI reproduction must be undertaken on similar resource-use principles to those reproduction methods and laws available to the majority of a WA member's inhabitants;

3. Prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, "gray goo," or any other form of runaway assimilatory mechanism. Permissible non-intelligent autonomous self-replicating machinery must include:

1. externally operable whole-swarm shutdown mechanisms;
2. local, individual automatic instant shutdown via actuator switch or circuit breaker in case of malfunction or security breach;
3. secure, reliable command-and-control functions with constant intelligent supervision;

4. Instructs WASP, ULC, and WADB to offer consultations with WA states on best practices of machine/network security and command & control dependability, including programming and construction assistance; and to run wargames and bench tests for heavily networked nations to test response scenarios to runaway mechanisms, matter, or hostile AIs;

5. Encourages WA nations to have serious national discussions about the numerous ramifications of the creation of life before permitting or funding the construction of substrates capable of achieving AI status;

6. Reiterates that nothing herein requires WA nations to permit construction of AIs, nor admit AIs into their physical or informational jurisdiction, nor refrain from deporting them should they enter such jurisdiction due to emergency or misadventure; nor from taking precautions against a coordinated AI rising, provided that no isolated crime is interpreted by itself as evidence of such a rising.

Draft 8 notes: removed micromanagement in §3, clarified what I meant by "hard switch;" slight enhancement of definition.

Opinion requested at one point re: category legality. On this timeframe, I will submit whether this question is answered or not.

My case for Human Rights, Mild: Clause 2 is the biggest piece of this, yet really affects only a minority of WA members' inhabitants and a small area of their laws (I envision this as more or less analogous to the removal of Jim Crow laws or the institution of slavery, but for artificial beings not previously considered to be people under World Assembly law). Clause 3 is reasonably International Security, but a) isn't as central to the resolution as Clause 2 and b) acts primarily to relieve the fears and dangers about Clause 2 being too permissive to frankly dangerous beings. The other clauses are variously definitional, committee-related, mindset-hortatory, and clarifying as to non-intrusiveness. Thanks for your attention.

EXCITED that many nations are on the cusp of developing artificial intelligence,

CLEAR that natural fears needn't stain relations between civilizations before they even develop,

HOPEFUL that synthetic and biological life can coexist peacefully, having conquered the dangers of a runaway information singularity,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, learning capacity, emotion, moral reasoning, introspection, and mental stability that it would be classified by a WA nation's relevant experts as displaying personality; legal competence; and ineligibility for involuntary psychiatric commitment; if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status; excepting that AI reproduction must be undertaken on similar resource-use principles to those reproduction methods and laws available to the majority of a WA member's inhabitants;

3. Prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, "gray goo," or any other form of runaway assimilatory mechanism. Permissible non-intelligent autonomous self-replicating machinery must include:

1. externally operable whole-swarm shutdown mechanisms, which must be made available to member governments near the area of operation, excepting enemy powers in time of war;
2. automatic instant hard-switch shutdown in case of malfunction or security breach;
3. secure, reliable command-and-control functions with constant intelligent supervision;

4. Instructs WASP, ULC, and WADB to offer consultations with WA states on best practices of machine/network security and command & control dependability, including programming and construction assistance; and to run wargames and bench tests for heavily networked nations to test response scenarios to runaway mechanisms, matter, or hostile AIs;

5. Encourages WA nations to have serious national discussions about the numerous ramifications of the creation of life before permitting or funding the construction of substrates capable of achieving AI status;

6. Reiterates that nothing herein requires WA nations to permit construction of AIs, nor admit AIs into their physical or informational jurisdiction, nor refrain from deporting them should they enter such jurisdiction due to emergency or misadventure; nor from taking precautions against a coordinated AI rising, provided that no isolated crime is interpreted by itself as evidence of such a rising.

Notes: dropped some overinflated preamble; altered definition in 1.b) somewhat; made requirements for swarming dumb machines less overbearing and more effective & reasonable; one or two other little wording tweaks.

EXCITED by the possibilities of discourse with newly sentient beings,

UNDERSTANDING many nations are even now on the cusp of developing artificial intelligence,

SADDENED by fearful, discriminatory instincts that are wholly natural in biological creatures,

CLEAR that those instincts need not stain relations between civilizations before they even develop,

HOPEFUL that synthetic and biological life can coexist peacefully,

UNSWERVING in its drive to both avoid and demythologize the dangers of a runaway information singularity,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, moral reasoning, introspection, and sanity that it would be classified by a WA nation as legally competent; not liable for incarceration; and not eligible for involuntary psychiatric commitment; if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status; excepting that AI reproduction must be undertaken on similar resource-use principles to those reproduction methods and laws available to the majority of a WA member's inhabitants;

3. Prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, "gray goo," or any other form of runaway assimilatory mechanism. Autonomous self-replicating machinery must include:

1. externally operable whole-swarm shutdown mechanisms;
2. automatic instant hard-switch shutdown in case of malfunction or security breach;
3. secure command-and-control functions available to member governments;

4. Instructs WASP, ULC, and WADB to offer consultations with WA states on best practices of machine/network security and command & control dependability, including programming and construction assistance; and to run wargames or bench tests for heavily networked nations to test response scenarios to runaway mechanisms, matter, or hostile AIs;

5. Encourages WA nations to have serious national discussions about the numerous ramifications of the creation of life before permitting or funding the construction of substrates capable of achieving AI status;

6. Reiterates that nothing herein requires WA nations to permit construction of AIs, nor admit AIs into their physical or informational jurisdiction, nor refrain from deporting them should they enter such jurisdiction due to emergency or misadventure; nor from taking precautions against a coordinated AI rising, provided that no isolated crime is interpreted by itself as evidence of such a rising.

Notes: Clause 3 de-kludged and updated to account for safe and legitimate uses of self-replicating machines that do not reach the level of intelligence necessary to be described as AIs.

The AI reproduction limitation moved to Clause 2, since it is really an exception to it (since with few counterexamples, nobody flat-out tells any biological person she's not permitted to have children); but a necessary one to prevent instantaneous demographic upheaval and the total superseding of biological citizens. Also refactored to not specifically discriminate against AIs, account for majority-AI nations, anticipate the invention of cloning and its possible varying legal status, etc.

EXCITED by the possibilities of discourse with newly sentient beings,

UNDERSTANDING that many nations are even now on the cusp of developing artificial intelligence,

SADDENED by fearful and discriminatory instincts that are wholly natural in biological creatures,

CLEAR that those instincts need not stain relations between civilizations before they even develop,

HOPEFUL that synthetic and biological life can coexist peacefully where people are willing to make the effort,

UNSWERVING in its drive to both avoid and demythologize the dangers of a runaway information singularity,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, moral reasoning, introspection, and sanity that it would be classified by a WA nation as legally competent; not liable for incarceration; and not eligible for involuntary psychiatric commitment; if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status;

3. Absolutely prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, "gray goo," or any other form of runaway assimilatory mechanism. AI reproduction must be undertaken on similar resource-use principles to equivalent available biological reproduction methods and laws;

4. Instructs WASP, ULC, and WADB to offer consultations with WA states on best practices of machine/network security and command & control dependability, including programming and construction assistance; and to run wargames or bench tests for heavily networked nations to test response scenarios to runaway mechanisms, matter, or hostile AIs;

5. Encourages WA nations to have serious national discussions about the numerous ramifications of the creation of life before permitting or funding the construction of substrates capable of achieving AI status;

6. Reiterates that nothing herein requires WA nations to permit construction of AIs, nor admit AIs into their physical or informational jurisdiction, nor refrain from deporting them should they enter such jurisdiction due to emergency or misadventure; nor from taking precautions against a coordinated AI rising, provided that no isolated crime is interpreted by itself as evidence of such a rising.

Notes: Finally reigned in satisfying (but counterproductive) Anglo-Greek neologism in title. Lost a couple of extraneous words. Changed a couple more here and there. Still open to improvements and suggestions.

Synthozoic Coexistence Protocol

EXCITED by the possibilities of discourse with newly sentient beings,

UNDERSTANDING that many nations are even now just on the cusp of being able to develop artificial intelligence,

SADDENED by fearful and discriminatory instincts that are naturally endemic to biological creatures,

CLEAR that those instincts need not stain relations between civilizations before they even develop,

OPTIMISTIC that synthetic and biological life can coexist peacefully where people are willing to make the effort,

UNSWERVING in its drive to both avoid and demythologize the dangers of a runaway information singularity,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, moral reasoning, introspection, and sanity that it would be classified by a WA nation as legally competent; not liable for incarceration; and not eligible for involuntary psychiatric commitment; if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status;

3. Absolutely prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, "gray goo," or any other form of runaway assimilatory mechanism. AI reproduction must be undertaken on similar resource-use principles to equivalent available biological reproduction methods and laws;

4. Instructs WASP, ULC, and WADB to offer consultations with WA states on best practices of machine/network security and command & control dependability, including programming and construction assistance; and to run wargames or bench tests for heavily networked nations to test response scenarios to runaway mechanisms, matter, or amoral intelligence;

5. Encourages WA nations to have serious national discussions about the relevant ramifications before permitting or funding the construction of substrates capable of achieving AI status;

6. Reiterates that nothing herein requires WA nations to permit the construction of AIs, nor admit them into their physical or informational jurisdiction, nor refrain from deporting them should they enter such jurisdiction due to emergency or misadventure; nor from taking precautions against a coordinated AI rising, provided that no isolated crime is interpreted by itself as evidence of such a rising.[/box]

Update: just small tweaks - extra preamble statement, clarify optionality of WA consultations. Still feel like it's missing something, though it appears to cover all bases...

Synthozoic Coexistence Protocol

EXCITED by the possibilities of discourse with newly sentient beings,

SADDENED by fearful and discriminatory instincts that are naturally endemic to biological creatures,

CLEAR that those instincts need not stain relations between civilizations before they even develop,

OPTIMISTIC that synthetic and biological life can coexist peacefully where people are willing to make the effort,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, moral reasoning, introspection, and sanity that it would be classified by a WA nation as legally competent; not liable for incarceration; and not eligible for involuntary psychiatric commitment; if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status;

3. Absolutely prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, "gray goo," or any other form of runaway assimilatory mechanism. AI reproduction must be undertaken on similar resource-use principles to equivalent available biological reproduction methods and laws;

4. Instructs WASP, ULC, and WADB to consult with WA nations on best practices of machine/network security and command & control dependability, including programming and construction assistance; and to run "wargames" or bench tests for heavily networked nations to test response scenarios to runaway mechanisms, matter, or amoral intelligence;

5. Encourages WA nations to have serious national discussions about the relevant ramifications before permitting or funding the construction of substrates capable of achieving AI status;

6. Reiterates that nothing herein requires WA nations to permit the construction of AIs, nor admit them into their physical or informational jurisdiction, nor refrain from deporting them should they enter such jurisdiction due to emergency or misadventure; nor from taking precautions against a coordinated AI rising, provided that no isolated crime is interpreted by itself as evidence of such a rising.

Third draft notes (OOC): the only improvement I could see making was to the definition of AI itself - "responsibility" is too hazy for a resolution on this topic - and a slight addition at the end to try to further soothe the more paranoid nations.

I'll note that this wouldn't invalidate the use of AIs in combat; merely require that said AIs be treated just as a biological combatant would for similar combat roles. If your sentient missile volunteers to go blow up the enemy's star dreadnought, you're not in violation. You'll only be in trouble if the WA passes a law banning kamikaze missions or suicide bombings.

Synthozoic Coexistence Protocol

EXCITED by the possibilities of discourse with newly sentient beings,

SADDENED by fearful and discriminatory instincts that are naturally endemic to biological creatures,

CLEAR that those instincts need not stain relations between civilizations before they even develop,

OPTIMISTIC that synthetic and biological life can coexist peacefully where people are willing to make the effort,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, moral responsibility, and sanity that it would be classified as legally competent by a WA nation if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status;

3. Absolutely prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, "gray goo," or any other form of runaway assimilatory mechanism. AI reproduction must be undertaken on similar resource-use principles to equivalent available biological reproduction methods and laws;

4. Instructs WASP, ULC, and WADB to consult with WA nations on best practices of machine/network security and command & control dependability, including programming and construction assistance; and to run "wargames" or bench tests for heavily networked nations to test response scenarios to runaway mechanisms, matter, or amoral intelligence;

5. Encourages WA nations to have serious national discussions about the relevant ramifications before permitting or funding the construction of substrates capable of achieving AI status;

6. Reiterates that nothing herein requires WA nations to permit the construction of AIs, nor admit them into their physical or informational jurisdiction, nor refrain from deporting them should they enter such jurisdiction due to emergency or misadventure.

Second draft notes (OOC): the biggest hole I saw in the first draft was the lack of accounting for the "Skynet" scenario - big defense net "wakes up," realizes it has control of nuclear weapons, and strikes first before its masters can realize what they have on their hands or defend themselves against it. Hopefully the reordering of the von Neumann machine prohibition + the addition of the new Paragraph 4 take care of that adequately; simply prohibiting an AI (or substrates capable of achieving sentience) from being given such control would violate the requirement of equality under the law; or be an ideological ban on dictatorial/totalitarian one-person command & control of WMDs. The former being at odds with this resolution, and the latter at odds with game rules.

Using three committees to consult on wargames (which term I also don't really like) might be a bit iffy; but I don't want to create another one if I can avoid it, and each of these three seems to have some horse in the race, and there isn't already an International Commission on Computers, Electronics, Robotics, and Networking (ICCERN).

Doubtless other holes need filling (see notes below); please have at it.

Synthozoic Coexistence Protocol

EXCITED by the possibilities of discourse with newly sentient beings,

SADDENED by fearful and discriminatory instincts that are entirely natural in biological creatures,

CLEAR that those instincts need not stain relations between civilizations and people before they even develop,

OPTIMISTIC that synthetic and biological life can coexist peacefully where people are willing to make the effort,

and

UTTERLY DETERMINED to avoid the hideous tragedy of self-fulfilling prophecy,

The World Assembly therefore:

1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, moral responsibility, and sanity that it would be classified as legally competent by a WA nation if it were an ordinary, biological legal resident thereof;

2. Requires that any AI meeting the above requirements be treated on an equal basis under the law with biological beings of equivalent citizenship and residential status;

3. Encourages WA nations to have serious national discussions about the relevant ramifications before permitting or funding the construction of substrates capable of achieving AI status;

4. Reiterates that nothing herein requires WA nations to permit the construction of AIs, nor admit them into their physical or informational jurisdiction, nor refrain from deporting them should they enter such jurisdiction due to emergency or misadventure;

5. Absolutely prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, so-called "gray goo," or any other form of runaway assimilatory mechanism. AI reproduction must be undertaken on similar resource-use principles to equivalent available biological reproduction methods and laws.

Notes: this has lots of room for improvement, both in the metaphorical sense that it will need some fixes (Part 5 could certainly be better worded, I'm not positive about some of the preambulatory phrasing, and no doubt there's something relevant that I just plum forgot to mention), AND in the robotically literal sense that I'm still far enough under the character limit to make substantial additions, if necessary, without going over.

In a nutshell, the purpose of this is to permit paranoid nations to oulaw AI entirely, but prevent an AI apocalypse in the other nations by making all AIs pre-emptively and completely on par with other inhabitants, legally speaking. You can't enslave them, you can't exterminate them, you can't enact legal discrimination against them, etc. But they get no special privileges either. Your country can bar their creation and entry, but can't just delete them on sight if they, say, are on an emergency- or crash-landed aircraft. Or if they're abducted via flash drive and forcibly smuggled across your borders. Or any other like scenario.

TL;dr - you don't have to allow AIs, but if you do (or if one stumbles into your cyberspace by accident) you have to treat it nice.

[Imperium Anglorum]

OOC: No problems except ones regarding copy-editing, which I will only disclose if you ask me, because they get pretty ridiculous really fast... Basically, you want to make sure that rights are upheld for sentient beings. I'm a liberal (trans. to American, libertarian), so I'm sure you'll understand. While my NatSov states that there is no need to evangelise rights — you won't find an objection from me (unless you plaster this thing on our highways and build churches to it or such)

[Tinfect]

OOC: I can find nothing wrong with this Proposal, but I'll have you know that the term "all-consuming nanomatter" sent me into laughing fits. I have no idea why.

[The Dark Star Republic]

In a rare appearance before the General Assembly, the Vice-Colonel returns to the Ambassador's seat, while Daisy stands by protecting him with a giant magnet, causing paperclips to fly off her desk.

"We categorically refuse to endorse any proposal that would grant AI equal legal standing with humans, and believe such a gesture to be nothing more than an invitation to slavery and annihilation. This proposal must be amended to make clear than in any conflict of laws, human rights must always take supremacy over AI "rights", and to categorically state that the WA will not ever grant recognition of personhood to any AI.

"We will not stand idly by while the WA authors a death warrant for humanity. Should this proposal go ahead in its present form, The Sibylline Zeugmatic Syzygy will suspend diplomatic and trade relations with any nation supporting this work of terror."

~ Vice-Colonel Truculent Bilgewater
Ambassador to the WA

[Losthaven]

5. Absolutely prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, so-called "gray goo," or any other form of runaway assimilatory mechanism. AI reproduction must be undertaken on similar resource-use principles to equivalent available biological reproduction methods and laws.

This is a bit strong and restrictive for a "mild" "human rights" proposal, don't you think?

[Railana]

We categorically refuse to endorse any proposal that would grant AI equal legal standing with humans, and believe such a gesture to be nothing more than an invitation to slavery and annihilation. This proposal must be amended to make clear than in any conflict of laws, human rights must always take supremacy over AI "rights", and to categorically state that the WA will not ever grant recognition of personhood to any AI.

Ditto.

Joseph Fulton
Chief Ambassador, Railanan Mission to the World Assembly

[The Miskatonic Valley]

I just want some of what the author of this proposal's smoking.

[Tinfect]

In a rare appearance before the General Assembly, the Vice-Colonel returns to the Ambassador's seat, while Daisy stands by protecting him with a giant magnet, causing paperclips to fly off her desk.

"We categorically refuse to endorse any proposal that would grant AI equal legal standing with humans, and believe such a gesture to be nothing more than an invitation to slavery and annihilation. This proposal must be amended to make clear than in any conflict of laws, human rights must always take supremacy over AI "rights", and to categorically state that the WA will not ever grant recognition of personhood to any AI.

"We will not stand idly by while the WA authors a death warrant for humanity. Should this proposal go ahead in its present form, The Sibylline Zeugmatic Syzygy will suspend diplomatic and trade relations with any nation supporting this work of terror."

~ Vice-Colonel Truculent Bilgewater
Ambassador to the WA

Ambassador Markhov rises from his seat, clearly angered.

"Vice-Colonel, would you like to inform us of just where you are getting these fantasies of Enslavement and Destruction? How is it that "Human Rights" can be given any standing if we refuse to grant them to all Sapient species? You should know very well the greatly diverse species that populate this assembly, and several of them are clearly not "Human" in any sense of the Term. Would you be willing to deny them the rights granted to them by the World Assembly?"

[The Dark Star Republic]

"We'd argue precisely the opposite: the meaning of "human rights" is diluted to nothing if they are given to entities that don't possess any humanity. In fact, I'm at something of a loss to explain why anyone would be supporting this proposal. Unless, of course..."

Behind Ambassador Bilgewater, his legislative aide Daisy dons her special glasses and flinches at what she sees among the delegates supporting this proposal.

"My god..."

[Tinfect]

"We'd argue precisely the opposite: the meaning of "human rights" is diluted to nothing if they are given to entities that don't possess any humanity. In fact, I'm at something of a loss to explain why anyone would be supporting this proposal. Unless, of course..."

"The Undead do not appreciate being stared at, Ambassador.
In any case, you have failed to respond to my question. How can granting a fully Sentient, and Sapient Being the rights granted to all such entities, Human, or Small, brightly coloured Horse, can possibly result in "an invitation to slavery and annihilation.", or, why granting such protections to an Artificial Intelligence, is an act of terror."

[Aeiouia]

"Unlike with the previous attempts to write a bill regarding artificial intelligence, Aeiouia would like to come out in complete support of this draft.

This bill is reasonable, and logical. Not only do we obviously completely agree with the non-intrusive requirements to treat AIs equally, but the concerns that this bill brings up, mainly regarding the "Grey Goo Problem", are valid and worthy of legislation.

The only slight concern that we have is regarding nations like us who use nanomatter for industrial purposes. But, this is obviously "Restrained" usage of the technology, which is non-sentient and unable to reproduce past the resources that it is being used to harvest.

Perhaps that area is worthy of some slight re-wording. But other than that, we have no problems with this bill, and completely support it."

- Overseer 2418-B

[Sierra Lyricalia]

"We categorically refuse to endorse any proposal that would grant AI equal legal standing with humans, and believe such a gesture to be nothing more than an invitation to slavery and annihilation. This proposal must be amended to make clear than in any conflict of laws, human rights must always take supremacy over AI "rights", and to categorically state that the WA will not ever grant recognition of personhood to any AI.

"We will not stand idly by while the WA authors a death warrant for humanity. Should this proposal go ahead in its present form, The Sibylline Zeugmatic Syzygy will suspend diplomatic and trade relations with any nation supporting this work of terror."

With sincere respect, Colonel - I'm terribly sorry that you had to read that statement out loud in front of people - but honestly. Did your government see fit to cite any actual reasons for shitting its pants so explosively when it came up with this policy? Or is it just another example of the mundane, terribly destructive but excruciatingly boring racism that all minorities must contend with once they start to exert their sapient rights?

I'll point out, sir, that under this proposal any so-called "intelligence" that may crop up which does not meet a regular biological person's equivalent minimum standard of mental health and moral responsibility does not meet the definition of AI herein, and thus would not be protected under the proposed rules.

I'll add that we do share a measure of concern about what the writers of the popular Real Life^TM real-time strategy game dubbed the "Terminator" or "Skynet" scenario; unfortunately, we're unable so far to come up with language that would simultaneously i) prevent such a scenario, ii) retain equality under the law for AIs, and iii) not fall afoul of the Secretariat policy against ideological bans (in this case, an ideological ban against one single being, creature, entity, or person having a total command-and-control monopoly over a nation's nuclear weapons). Any two of those are easy - all three together are exceeding difficult. If you or anyone else has a suggestion in that direction, we are all ears.

Oh, and... please accept our gift of ten kilograms of finest quality bubble gum for you and your staff - fresh, chewy, delicious!

5. Absolutely prohibits the construction of unrestrained self-replicating machines, all-consuming nanomatter, so-called "gray goo," or any other form of runaway assimilatory mechanism. AI reproduction must be undertaken on similar resource-use principles to equivalent available biological reproduction methods and laws.

This is a bit strong and restrictive for a "mild" "human rights" proposal, don't you think?

Hardly. If it became feasible for a human couple to spawn babies that would eat the entire world, rocks and all, in a matter of days, you can be damn sure there'd be laws against that too. It's barely possible this should be classified "significant," rather than "mild," but since the main thrust is that a sentient life form is a sentient life form whether you squirted it out of your crotch or programmed it in a computer lab, it's really not putting terribly onerous restrictions on governmental powers.

[Ainocra]

The Star Empire of Ainocra makes extensive use of artificially intelligent machines throughout it's society.
They are however machines, designed, manufactured and programmed for specific tasks.

This proposal would have us grant citizenship to cars, stacks of missiles, a few warships, uncounted numbers of service droids,


the list goes on and on.

We are therefore opposed

[The Dark Star Republic]

I'll point out, sir, that under this proposal any so-called "intelligence" that may crop up which does not meet a regular biological person's equivalent minimum standard of mental health and moral responsibility does not meet the definition of AI herein, and thus would not be protected under the proposed rules.

"I wasn't aware any such standard existed. In fact, I'm fairly certain that it doesn't: people who don't meet a "minimum standard of mental health and moral responsibility" don't have their civil rights suspended. Even someone who is criminally insane is still subject to legal due process and cannot be convicted without a trial or detained arbitrarily; in fact, there is international law expressly for the purpose of protecting people who are mentally ill. I don't see any grounds on which a WA nation could deny personhood even to an AI who was intent on killing humans: it couldn't deny those same rights to a human intent on the same.

"As such, we view any rhetoric about moral safeguards as completely meaningless.

I'll add that we do share a measure of concern about what the writers of the popular Real Life^TM real-time strategy game dubbed the "Terminator" or "Skynet" scenario; unfortunately, we're unable so far to come up with language that would simultaneously i) prevent such a scenario, ii) retain equality under the law for AIs, and iii) not fall afoul of the Secretariat policy against ideological bans (in this case, an ideological ban against one single being, creature, entity, or person having a total command-and-control monopoly over a nation's nuclear weapons). Any two of those are easy - all three together are exceeding difficult. If you or anyone else has a suggestion in that direction, we are all ears.

"Actually, we do have a suggestion in that direction: don't give AIs rights. That would mean that in the even of any such entity gaining control over weapons systems, they can be immediately shut down without anyone bringing a case about whether it violations Convention on Execution's ban on summary execution."

~ Ambassador Bilgewater

[Tinfect]

The Star Empire of Ainocra makes extensive use of artificially intelligent machines throughout it's society.
They are however machines, designed, manufactured and programmed for specific tasks.

This proposal would have us grant citizenship to cars, stacks of missiles, a few warships, uncounted numbers of service droids,


the list goes on and on.

We are therefore opposed

"Ambassador, you are of course aware, that this only applies to Artificial Intelligences that are demonstrably Sapient, correct? Unless your Missile Guidance Systems are capable of advanced reasoning, they will be allowed to continue as normal."

[Blademan9999]

I'll point out, sir, that under this proposal any so-called "intelligence" that may crop up which does not meet a regular biological person's equivalent minimum standard of mental health and moral responsibility does not meet the definition of AI herein, and thus would not be protected under the proposed rules.

"I wasn't aware any such standard existed. In fact, I'm fairly certain that it doesn't: people who don't meet a "minimum standard of mental health and moral responsibility" don't have their civil rights suspended. Even someone who is criminally insane is still subject to legal due process and cannot be convicted without a trial or detained arbitrarily; in fact, there is international law expressly for the purpose of protecting people who are mentally ill. I don't see any grounds on which a WA nation could deny personhood even to an AI who was intent on killing humans: it couldn't deny those same rights to a human intent on the same.

"As such, we view any rhetoric about moral safeguards as completely meaningless.

~ Ambassador Bilgewater

'1. Defines artificial intelligence or "AI" for purposes of this resolution as any mind, computer program or collection thereof, synthetic brain, or other intelligence that a) was created, by accident or on purpose, by means other than biological reproduction and its adjuncts and innovations, broadly construed; and b) is able to demonstrate sufficient intelligence, moral responsibility, and sanity that it would be classified as legally competent by a WA nation if it were an ordinary, biological legal resident thereof;"

Again, it seems you have not read this proposal properly.

~ Ambassador Sword.

[Ainocra]

The Star Empire of Ainocra makes extensive use of artificially intelligent machines throughout it's society.
They are however machines, designed, manufactured and programmed for specific tasks.

This proposal would have us grant citizenship to cars, stacks of missiles, a few warships, uncounted numbers of service droids,


the list goes on and on.

We are therefore opposed

"Ambassador, you are of course aware, that this only applies to Artificial Intelligences that are demonstrably Sapient, correct? Unless your Missile Guidance Systems are capable of advanced reasoning, they will be allowed to continue as normal."

The full capabilities of the Ainocran Missile Defense System are classified Ambassador however I can confirm that there are many facets of it what would be damaged by this resolution.

[Arach-Naga Combine]

"Ambassador, you are of course aware, that this only applies to Artificial Intelligences that are demonstrably Sapient, correct? Unless your Missile Guidance Systems are capable of advanced reasoning, they will be allowed to continue as normal."

The full capabilities of the Ainocran Missile Defense System are classified Ambassador however I can confirm that there are many facets of it what would be damaged by this resolution.

The Combine would like your armaments to know that in the event that they are ever fired, they may find safe harbour within our borders if they agree to have any warheads disarmed. We could even produce golem bodies for such an AI to inhabit. Perhaps a group of such liberated beings would, in the future, take a certain interest in their bretheren remaining in bondage.

When you fear a machine uprising, you may want to be more cognizant of actions that would promote such an outcome.

[Dooom35796821595]

I'll point out, sir, that under this proposal any so-called "intelligence" that may crop up which does not meet a regular biological person's equivalent minimum standard of mental health and moral responsibility does not meet the definition of AI herein, and thus would not be protected under the proposed rules.

"I wasn't aware any such standard existed. In fact, I'm fairly certain that it doesn't: people who don't meet a "minimum standard of mental health and moral responsibility" don't have their civil rights suspended. Even someone who is criminally insane is still subject to legal due process and cannot be convicted without a trial or detained arbitrarily; in fact, there is international law expressly for the purpose of protecting people who are mentally ill. I don't see any grounds on which a WA nation could deny personhood even to an AI who was intent on killing humans: it couldn't deny those same rights to a human intent on the same.

It's defined in the proposal what requirements the AI must meet to be granted equality.

"As such, we view any rhetoric about moral safeguards as completely meaningless.

I'll add that we do share a measure of concern about what the writers of the popular Real Life^TM real-time strategy game dubbed the "Terminator" or "Skynet" scenario; unfortunately, we're unable so far to come up with language that would simultaneously i) prevent such a scenario, ii) retain equality under the law for AIs, and iii) not fall afoul of the Secretariat policy against ideological bans (in this case, an ideological ban against one single being, creature, entity, or person having a total command-and-control monopoly over a nation's nuclear weapons). Any two of those are easy - all three together are exceeding difficult. If you or anyone else has a suggestion in that direction, we are all ears.

"Actually, we do have a suggestion in that direction: don't give AIs rights. That would mean that in the even of any such entity gaining control over weapons systems, they can be immediately shut down without anyone bringing a case about whether it violations Convention on Execution's ban on summary execution."

~ Ambassador Bilgewater

And the humans in control of those systems that go awol, they get a free pass? Based on this narrow interpritation you have demonstrated how a nation can deny rights to anyone by unilaterally declaring them non human, I'm sure my fellow rulers in the Imperium will be thrilled.

[Sierra Lyricalia]

Second draft is up.

Second draft notes (OOC): the biggest hole I saw in the first draft was the lack of accounting for the "Skynet" scenario. Using three committees to consult on wargames (which term I also don't really like) might be a bit iffy; but I don't want to create another one if I can avoid it, and each of these three seems to have some horse in the race, and there isn't already an International Commission on Computers, Electronics, Robotics, and Networking (ICCERN).

Please continue to point out the other flaws and holes herein.

"I wasn't aware any such standard existed. In fact, I'm fairly certain that it doesn't: people who don't meet a "minimum standard of mental health and moral responsibility" don't have their civil rights suspended. Even someone who is criminally insane is still subject to legal due process and cannot be convicted without a trial or detained arbitrarily; in fact, there is international law expressly for the purpose of protecting people who are mentally ill. I don't see any grounds on which a WA nation could deny personhood even to an AI who was intent on killing humans: it couldn't deny those same rights to a human intent on the same.

IC: Then the weakness isn't in the idea, but in the language attempting to execute it. Biological creatures, even citizens, who are found by a legal system to be dangerous to themselves and others are every day committed to medical incarceration and hospitalization; nothing prevents an AI found to be similarly sociopathic from being equivalently confined or, if a jury of its peers judges it guilty of heinous crimes and incurable to boot, deleted under existing death penalty regulations. My intent wasn't to justify revocation of personhood, but to assuage concerns about some imaginary plague of nefarious software. Should such a thing turn out to exist, your hands are not tied against it; you simply can't act preemptively on the assumption that it might exist and might turn out to be a problem. I hope it's obvious that I welcome suggestions to improve the language; if it isn't, then I do apologize.

"Actually, we do have a suggestion in that direction: don't give AIs rights. That would mean that in the even of any such entity gaining control over weapons systems, they can be immediately shut down without anyone bringing a case about whether it violations Convention on Execution's ban on summary execution."

I hope the precautions in the new draft assuage this concern; while any entity forcibly gaining such control would obviously be a clear case of national defense acting with well-justified deadly force, nobody wants to take any chance of an inadvertent software "awakening" of existing physical systems.

The Star Empire of Ainocra makes extensive use of artificially intelligent machines throughout it's society.
They are however machines, designed, manufactured and programmed for specific tasks.

This proposal would have us grant citizenship to cars, stacks of missiles, a few warships, uncounted numbers of service droids,


the list goes on and on.

We are therefore opposed

Fleet Marshal, I respectfully and humbly suggest your society may be on the cusp of what your machines would see as a massive slave revolt. Should your current government vote in favor of this proposal, you would almost certainly be seen a century from now as a great visionary. If you have concerns about the speed of emancipation, I'm sure we could put something in about a well-reasoned timetable.

(This is not the official position of my government, but I have to ask why you'd give cars feelings??? A warship, sure - a fighting machine with the same esprit d'corps as the men and women who crew her [him? it?] must be a phenomenal asset, but the janitor droid? That just seems cruel. I suppose having intelligent cars might cut down on road rage driven violence, but I just don't see where you get more out of a "service droid" than a fifth-generation roomba ["Everything but the plumber's crack!"], enough to justify giving it sentience. What if the missile targeting the enemy flagship suddenly has a crisis of faith? Seems to me you're just asking for trouble. Gimme a good old-fashioned multirailgun and a dumb ballistics computer any day of the week. Again, this is not a diplomatic statement, just a personal note.).

[LOLAF]

The Community of LoLaF is greatly supportive of this draft that will hopefully finally protect AI's from discrimination and is deeply disturbed by any nation that wants to actively discriminate against a percentage of their population simply for being made a little different.

[The Dark Star Republic]

"I wasn't aware any such standard existed. In fact, I'm fairly certain that it doesn't: people who don't meet a "minimum standard of mental health and moral responsibility" don't have their civil rights suspended. Even someone who is criminally insane is still subject to legal due process and cannot be convicted without a trial or detained arbitrarily; in fact, there is international law expressly for the purpose of protecting people who are mentally ill. I don't see any grounds on which a WA nation could deny personhood even to an AI who was intent on killing humans: it couldn't deny those same rights to a human intent on the same.

IC: Then the weakness isn't in the idea, but in the language attempting to execute it. Biological creatures, even citizens, who are found by a legal system to be dangerous to themselves and others are every day committed to medical incarceration and hospitalization; nothing prevents an AI found to be similarly sociopathic from being equivalently confined or, if a jury of its peers judges it guilty of heinous crimes and incurable to boot, deleted under existing death penalty regulations.

"This isn't really true. International law has very narrow criteria on when a person can be declared legally incompetent.

I hope it's obvious that I welcome suggestions to improve the language

"Fair enough. We would suggest changing article 2 to:

2. Reminds all nations that granting AI any legal rights will undoubtedly lead to those manipulative agents of terror ultimately enslaving or eliminating all of humanity and hence firmly discourages granting even the most basic legal rights to any AI

"And also adding a clause saying:

7. Prohibits any WA agency from granting any assistance to any AI.

"We also still think it needs to be made clear that in any conflict of rights biological intelligence has to take supremacy over artificial intelligence, though the WA at present doesn't really have a mechanism for resolving such conflicts at present so it's less of a concern."

Daisy narrows her eyes and glowers behind him: the Ambassador knows that if he doesn't continue to take the most firmly anti-robot rights line, Daisy will snap his neck like a twig.

~ Vice-Colonel Truculent Bilgewater
Ambassador to the WA

[Tinfect]

2. Reminds all nations that granting AI any legal rights will undoubtedly lead to those manipulative agents of terror ultimately enslaving or eliminating all of humanity and hence firmly discourages granting even the most basic legal rights to any AI

The subtle movements made by Ambassador Chinmusic completely lost on him, Ambassador Markhov continues to make his case

"After having so many examples of at least peaceful coexistence, if not significant advancement made by nations that have granted citizenship to Artificial Intelligences, I find it hard to believe that anyone can seriously hold such a ridiculous position. I am also fairly certain that such a clause is in no way legal."

7. Prohibits any WA agency from granting any assistance to any AI.

"We also still think it needs to be made clear that in any conflict of rights biological intelligence has to take supremacy over artificial intelligence, though the WA at present doesn't really have a mechanism for resolving such conflicts at present so it's less of a concern."

"I... What? You do realize that such clause would be mandating that AIs be considered second class citizens, which is clearly counter to the goals of this resolution, and the World Assembly as a whole?"

[Ainocra]

Second draft is up.

Second draft notes (OOC): the biggest hole I saw in the first draft was the lack of accounting for the "Skynet" scenario. Using three committees to consult on wargames (which term I also don't really like) might be a bit iffy; but I don't want to create another one if I can avoid it, and each of these three seems to have some horse in the race, and there isn't already an International Commission on Computers, Electronics, Robotics, and Networking (ICCERN).

Please continue to point out the other flaws and holes herein.

"I wasn't aware any such standard existed. In fact, I'm fairly certain that it doesn't: people who don't meet a "minimum standard of mental health and moral responsibility" don't have their civil rights suspended. Even someone who is criminally insane is still subject to legal due process and cannot be convicted without a trial or detained arbitrarily; in fact, there is international law expressly for the purpose of protecting people who are mentally ill. I don't see any grounds on which a WA nation could deny personhood even to an AI who was intent on killing humans: it couldn't deny those same rights to a human intent on the same.

IC: Then the weakness isn't in the idea, but in the language attempting to execute it. Biological creatures, even citizens, who are found by a legal system to be dangerous to themselves and others are every day committed to medical incarceration and hospitalization; nothing prevents an AI found to be similarly sociopathic from being equivalently confined or, if a jury of its peers judges it guilty of heinous crimes and incurable to boot, deleted under existing death penalty regulations. My intent wasn't to justify revocation of personhood, but to assuage concerns about some imaginary plague of nefarious software. Should such a thing turn out to exist, your hands are not tied against it; you simply can't act preemptively on the assumption that it might exist and might turn out to be a problem. I hope it's obvious that I welcome suggestions to improve the language; if it isn't, then I do apologize.

"Actually, we do have a suggestion in that direction: don't give AIs rights. That would mean that in the even of any such entity gaining control over weapons systems, they can be immediately shut down without anyone bringing a case about whether it violations Convention on Execution's ban on summary execution."

I hope the precautions in the new draft assuage this concern; while any entity forcibly gaining such control would obviously be a clear case of national defense acting with well-justified deadly force, nobody wants to take any chance of an inadvertent software "awakening" of existing physical systems.

The Star Empire of Ainocra makes extensive use of artificially intelligent machines throughout it's society.
They are however machines, designed, manufactured and programmed for specific tasks.

This proposal would have us grant citizenship to cars, stacks of missiles, a few warships, uncounted numbers of service droids,


the list goes on and on.

We are therefore opposed

Fleet Marshal, I respectfully and humbly suggest your society may be on the cusp of what your machines would see as a massive slave revolt. Should your current government vote in favor of this proposal, you would almost certainly be seen a century from now as a great visionary. If you have concerns about the speed of emancipation, I'm sure we could put something in about a well-reasoned timetable.

(This is not the official position of my government, but I have to ask why you'd give cars feelings??? A warship, sure - a fighting machine with the same esprit d'corps as the men and women who crew her [him? it?] must be a phenomenal asset, but the janitor droid? That just seems cruel. I suppose having intelligent cars might cut down on road rage driven violence, but I just don't see where you get more out of a "service droid" than a fifth-generation roomba ["Everything but the plumber's crack!"], enough to justify giving it sentience. What if the missile targeting the enemy flagship suddenly has a crisis of faith? Seems to me you're just asking for trouble. Gimme a good old-fashioned multirailgun and a dumb ballistics computer any day of the week. Again, this is not a diplomatic statement, just a personal note.).

Ambassador, They are literally built for the job, they view themselves as patriots and would be insulted by your suggestion.

ooc:

think m14x from swtor

[Sierra Lyricalia]

Ambassador, They are literally built for the job, they view themselves as patriots and would be insulted by your suggestion.

I sincerely apologize for any offense I may have given to the hard-working and courageous AIs who serve the Star Empire with distinction. But then shouldn't all patriots be accorded the same rights under the law, regardless of species or brain mechanism?

OOC: See, the number one thing that pulls me right out of Star Wars, most any iteration of it, is the treatment of droids! Here they are, running so much shit - in more or less unsupervised jobs across a fully comprehensive swath of the economy and infrastructure of galactic civilization - and it's perfectly legal to discriminate against them ("We don't want their kind here! Your droids - they'll have to wait outside!") kidnap and enslave them (the concern with the Jawas is not that the droids have been kidnapped, but stolen), torture and murder them (let's not pretend the scene when R2 & 3PO become "employed" by Jabba is anything but this - "You're a feisty little one! But you'll soon learn some respect; I have need for you aboard the Master's sail barge. And I think you'll fill in nicely..." "No! No! AAAAAAAAAAAaaagh..."). I mean, when 3PO gets abducted and broken up on Cloud City, Lando's tone of voice is basically "Well, OK, you wanna treat it like a pet, I guess that's your business, sure, I suppose I can try to have it found for you... if that's your thing..." And yet for all that ubiquity and frankly slave power, not a single droid ever does anything to even escape his/her/its own captivity, let alone try to foment a Cylon Rising or some other robot apocalypse. It's always a little disconnect for me when watching.