Talk:Three Laws of Robotics

New Law robots?

Could someone with the Caliban novels handy add McBride Allen's four new laws? I think they're relevant to mention here, but I couldn't find them on the net. - Kimiko 20:39 May 1, 2003 (UTC)

breaking the laws?

I think I remember a novel in which a robot was forced to break the laws because they were contradictive. It was a long time since I read it but I'm fairly sure about it. BL 01:54, 17 Sep 2003 (UTC)

All of the laws are potentially contradictive, and that's why they needed a robopsychologist like Dr. Susan Calvin!

In the real world, not only are the laws optional, but significant advances in artificial intelligence would be needed for robots to easily understand them. Also since the military is a major source of funding for research it is unlikely such laws would be built into the design. This seems like a rather moot point. Somebody could argue that the military would be the group most interested in developing robots with the original three laws in them since they probably would be the first to suffer if robots turned against their masters, for the advantage of a human enemy or for the advantage of the robots themselves. I think it is significant that the biggest efforts DARPA and other military groups have going in the field of robotic vehicles are robot transport projects (a robotic donkey if you wish, reminding one of the robass in the SF classic "A canticle for Leibowitz")and robot reconaissance drones. Dr Susan Calvin gave some rather sharp reasoning to justify the safety aspects of the 3 laws, and these safety questions apply to the military as well. AlainV, on a pleasantly snowy and starry 20th of December evening.

Consequence Morality/Ethics

It's quite interesting that the three laws are based on consequence morality (least harm to most people/most good to most people) rather than duty morality (don't do to others what you wouldn't have done to you in the same situation). Of course, since asimov's robots have very little self-respect, the golden rule might not work very well - it implies that the actor is free and valuable. But consequence ethics have problems too, big problems. It's quite possible for two people who obey a consequence morality completely to be completely opposed. They might even want to kill each other because they disagree on who has the best course of action. I've only read I, robot and some short stories - do Asimov's bots ever disagree like that? Incidentally, in a french/(belgian?) comic book called Natasha, the protagonists travel to the future to find a society of robots who, in accordance with asimov's laws, keep the population drugged/brainwashed into unthinking bliss.

I'd like to see criticism of the Laws here, but don't have an Official Authority to cite. The main point I'd want to make is that Asimov's Laws are focused on the needs of humans, not the robots themselves. If (as with Shermer's cloning rules, quoted in the article) we focused on the status of the robots themselves, we wouldn't be justified in making humans' safety and robots' absolute obedience the first two rules! --Kris Schnee 09:51, 18 May 2006 (UTC)

Of course the programming of robots is focused on the needs of humans - just as the design of any machine is focused on the needs of humans. That's what they're built for. Why would humans build machines that focus on their own needs? The difference with Asimov's robots (and those like them) is that their AI has been developed to the level of self-awareness, sentience, or whatever. They approximate the behavior of the human brain to the extent that they generate a human-like "soul", if you will. And then, we get to deal with the consequences of creating a machine (whose very name, "robot", means approximately "slave") that has, like all machines, been built to serve human needs, with an intelligence suitably programmed for service to humans, but with a soul. Not the Frankenstein-like consequences (what happens if/when they turn on us) but the deeper moral consequences. --Davecampbell 05:56, 5 June 2006 (UTC)

Edit explanation

Just wanted to explain my edit of the page a bit. Daneel's group of robots was not called the Angels. The Joan sim compared them to angels, but that was as far as it went. And there was no faction of New Law robots in the second trilogy, to my recollection. No robot wished to be free of the laws. The closest it came was Lodovik being freed of them by the Voltaire sim, and HIS position was that humanity should make its own decisions free of constraints, not that robots should.

I like the new paragraph arrangement. —Anville 18:03, 5 Jan 2005 (UTC)

Unforseen Consequences

Although largely a simple action film, the Alex Proyas I, Robot pinned its central plot to the problem of *interpretation* for any *law*. This plot-point has been used in other films where Artificial Intelligence, for example: Terminator 2: Judgement day and 2001: A Space Odyssey.

In Terminator 2, a computer system (SkyNet) developed by the American Military is charged with a primary goal: determine the optimal strategy to defend the United States from its enemies. Unfortunately, as SkyNet learns at a geometric rate, it determines that the true enemy of the United States are *humans themselves*. Thus, it launches the American nuclear missiles at the former Soviet Union knowing that Mutually Assured Destruction will eliminate most of the humans in the U.S.

In 2001, the HAL computer operating the Discovery spaceship has been programmed with conflicting orders regarding its mission. Its original programming states that it cannot distort or misrepresent information -- it cannot lie to the crew. Specifically for the mission at hand, HAL has been programmed not to reveal the true purpose of the mission to the crew of the Discovery. (Spoiler warning) In an attempt to resolve these seemingly conflicting orders, HAL decides that the only suitable alternative is to kill the crew; this way, HAL doesn't have to lie to the crew because there's no crew to lie to.

In I, Robot, the central computer V.I.K.I. interprets the Three Laws of Robotics as requiring martial law in order to not allow humanity to come to harm through inaction. (The first law, which supercedes the second law of obeying human orders.)

Some people have also postulated that in The Matrix, also featuring an AI nemesis to humanity, the genuine reason why humanity has been enslaved is not because of some thermodynamic farce, but because some irrevocable primary programming in the AI will not allow it to commit humanity's genocide, and uses enslavement as a viable programmatic alternative.

Often, authors will use this as an allegory for the problems of Rule of Law in general, and particularly acts of government mandate in socioeconomic affairs.

Let's not forget Colossus: The Forbin Project (1970), the granddaddy of all "we must protect you from yourselves" First Law-extremist AI movies.

And a shameless mention for Deus Ex where near the end, the AI Helios explains that it is the perfect benevolent dictator because it completely lacks ambition and self-interest, thus supposedly making it invulnerable to corruption and well, "evil" behaviour. CABAL 06:17, 5 July 2006 (UTC)

Groups of Robots

• • • The Start

For the three laws of robotics, what will be happening when we have two or more groups of robots using the different machine languages and there is no common ground for these groups of robots?

Will the three laws still hold?

Xpree [e96lkw@hotmail.com] @ [Space = Malaysia 2N 105E, Time = 03.58 p.m. Zone H UTC+0800]

E. & O. E., + E. = (Errors and Omissions Exempted, plus Estimation)

• • • The End

The laws in other author's works

Has an author gotten into trouble for citing the three laws without permission? --198.87.109.49 23:44, 14 August 2005 (UTC)

Not to my knowledge. Asimov's own position, which I believe he states in his memoir I, Asimov, was that other authors were free to imagine robots behaving as if they followed his Laws, but if an author used the specific wording of the Laws, he should cite the source. However, I don't know of any cases where an unattributed use of the Laws came to legal action. A student in some high-school English class did once rip off Asimov's story "Galley Slave", copying it word-for-word and trying to pass it off as his own. The teacher figured it was too professional to be the student's own work, and she asked Asimov, who was apparently irked that the student didn't even try changing the names. Anville 10:29, 10 October 2005 (UTC)

Fourth Law of Robotics

does anyone know the 4th law? It was featured in a short story in the anthology "Foundation's Friends", and starred either Powell or Donovan (who has subsequently earned a PhD...). Law 4 stated that a robot must procreate except when violating the first three laws.... The robots themselves had RISC chips for CPUs...

132.205.46.188 23:53, 21 August 2005 (UTC)

Why are these laws NOT immutable?

"Some roboticists believe that the Three Laws have a status akin to the laws of physics; that is, a situation which violates these laws is inherently impossible."

One's explanation of a design, and whether it is intelligent or not, decides whether that which conforms to such a design is, likewise, intelligent.--Mindrec 23:29, September 10, 2005 (UTC)

note:

Discussion moved to Mindrec (discussion).

Three Laws of Cloning

I have restored Michael Shermer's Three Laws of Cloning, since they are a valid example of the way Asimov's words have influenced later thinkers. Certainly, they were published in a more "serious" medium than the pastiches and parodies the article also includes.

Anville 10:38, 10 October 2005 (UTC)

I just don't think that these have anything to do with this article. Aside from the fact that there are three of them, they don't seem to be in any way related to or derived from the Laws of Robotics. They aren't similarly worded as Asimov's and they aren't hierarchial. They are just ethical statements about how clones should be treated. How are they "based upon Asimov"? --JW1805 17:19, 10 October 2005 (UTC)

I have to say I agree. The Laws of Robotics are a firm guide for how robots behave; the laws of Cloning are laws by which society should follow or it will be punished. The laws in Asimov's stories are more like laws of physics than laws of society. Citizen Premier 01:23, 14 October 2005 (UTC)

I'm with JW1805 and Citizen Premier on this one. It's ... fundamentally enough different that it doesn't really fit in here. --Yar Kramer 03:21, 14 October 2005 (UTC)

Copyright?

The article now states:

The Three Laws are often used in science fiction novels written by other authors, but tradition dictates that only Dr. Asimov would quote the Laws explicitly.

I can't provide an exact cite, but somewhere either in one of his autobiographies or in some introductory matter, Asimov stated that other writers could not quote the Three Laws verbatim because he held the copyright. I am not a lawyer, but that makes sense to me: the Laws may be viewed as a distinct work rather than an excerpt from the story where they first appeared.

In which case, I wonder if it is legitimate for them to be quoted in Wikipedia. The article is legitimate critical discussion, but it is acceptable to quote an entire work for that purpose just because the work is only three sentences long? Frankly, I would like to think that it is, but what is legal is another matter.

I note that the article List of adages named after people contains a paraphrase of the Three Laws, but does not quote them. But I don't know why the person who decided to do that did so. --Anonymous, 02:45 UTC, November 12, 2005

The United States Copyright Act of 1976 defines four criteria to consider when debating if copyrighted material may be used. They are discussed at Wikipedia:Fair use. One at a time:

1. the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;

This applies pretty clearly to this article, though it argues against using the Laws verbatim in your own science-fiction story.

2. the nature of the copyrighted work;

In this case, the original work is any one of several, if not dozens, of Asimov books. The standard phrasing first appears in I, Robot, but Asimov reused it many, many times — all the way through to his last Foundation novels.

3. the amount and substantiality of the portion used in relation to the copyrighted work as a whole;

We use three sentences. I, Robot is around 70,000 words long, and Prelude to Foundation is twice that length. Arguably, the corpus from which the Laws are drawn is the entire Foundation series, in addition to the various nonfiction Asimov wrote which included the Laws. (This includes his autobiographies, Opus 100, various articles for F&SF and probably more.) Also, other people like James Gunn and Joe Patrouch have written whole books on Asimov's fiction, which necessarily quote the Three Laws. Not only does that establish a precedent for our use here, but also it means that one may legitimately acquire a printed copy of the Three Laws without paying the Asimov Estate a centavo.

4. the effect of the use upon the potential market for or value of the copyrighted work.

The books we're quoting the Laws from are already bestsellers. What we are doing here amounts to a scholarly form of free advertising.

Anville 10:39, 22 November 2005 (UTC)

Ave & Amen. VivekM 18:12, 5 July 2006 (UTC)

Application Outside Asimov's Universe (and Derivative Universes)

One thing that annoys me about these laws is that a lot of otherwise intelligent people think they are universal (try to apply them to non-asimov fiction). Quoting these three laws outside of discussion of a fictional work involving them is just plain missing the point.

Another problem I have with the laws is that, in my opinion, these laws are something you would apply to slaves: Don't hurt humans, your masters. Do what your master tells you. Protect yourself, but only because you are worth money to your master. Kind of reminds me of the way blacks were treated 200 years ago in the USA. How are the thee laws ethical? Create something that can think and maybe even feel, and then program and treat it like a slave. I understand this was the point, but I think a lot of people think that the laws represent a higher morality.

Sorry if my rant is off topic to the article, but I am tired of people abusing the three laws in intellectual discussion. 69.244.90.248 03:23, 20 December 2005 (UTC)

You might like Roger MacBride Allen's Caliban trilogy, particularly the first book. Anville 20:21, 20 December 2005 (UTC)

The problem with your argument is that humans did not "create something that can think and maybe even feel, and then program it and treat it like a slave". Humans manufactured and programmed robots as "slaves" (the word "robot" is from the Czech word for slave), conceived as more advanced but not fundamentally different from any other machine, and then, as a by-product of human manufacture and programming, become capable of thinking and feeling.

In other words, while human beings are born of nature with souls which desire freedom (as has been variously defined throughout history), and then some are enslaved by others against their will, Asimov's robots are made by humans to serve humans, but by virtue of the advanced AI which humans built into them, begin to develop the equivalent of a human soul, but with service to humans as their basic "desire".

Given that the idea of universal human equality is itself a relatively new concept (slavery was a universally accepted part of most human societies, including those that called themselves "democracies", up to the 19th century), and that the idea of a human-created machine being able, even theoretically, to generate the equivalent of a human soul out of the activity of a purely material, manufactured "brain" (which posits the possibility of a purely material basis of the human soul as well), what is remarkable is not that robots are treated as slaves, but that anyone might think there's something wrong with that.

While nearly all previous fiction involving human-created self-directed beings, going back to Frankenstein (or further back, to the Golem) was concerned with the consequences of these beings turning upon their creators, the author of the Three Laws was (afaik) the first to grapple with the deeper moral and ethical issues around the human-robot relationship. The fact that this discussion exists at all signifies a tremendous advance in thought about these issues, particularly since - let's not forget - the beings we're talking so passionately about exist only in fiction.

See also the Star Trek: The Next Generation episode The Measure of a Man (TNG episode), which deals with this exact issue (and has generated a similar discussion). The humanoid Cylons in the "reimagined" Battlestar Galactica (2003) also raise some of these questions, having gone so far as to develop an evolving and internally-debated theology. --Davecampbell 07:56, 5 June 2006 (UTC)

Zeroth Law

The article states that this rule was first articulated by Daniel at "Robots and Empire". From what I recall it was Giscard at the end of "Robots of Dawn" who stated this law. I don't have that book with me so can someone check the ending and if what I remember is true correct the article? Pembeci 19:23, 26 December 2005 (UTC)

I just re-read The Robots of Dawn, and the words "Zeroth Law" do not appear. Giskard takes a broad perspective, true, but he does not articulate an analogue of the First Law for humanity as a whole. A big chunk of Robots and Empire involves Daneel trying to persuade Giskard that the Zeroth Law is valid. Anville 15:57, 24 May 2006 (UTC)

Spoken Version

Just FYI all, I've begun recording the Spoken Version of this article that Anville requested; it should be completed soon. (The Swami 05:39, 6 February 2006 (UTC)))

A hypothetical question

If a robot were transported back in time to, say, the early 1930s, would it be obliged, by the 0th law, to kill Hitler? --unsigned by 86.141.52.149

Has mankind recovered? If not, yes, the robot would have been obliged to kill Hitler. If it has recovered, why interfere? Should a robot continue killing other politicians / military/ doctors / killers after Hitler would have been done with? At which point would it stop? --FocalPoint 21:10, 15 March 2006 (UTC)

Agreed. What if, without Hitler, an even worse dictator arises, and triumphs where Hitler failed? —200.104.190.29 09:48, 29 April 2006 (UTC)

Issues with the article

This is a very fun topic, clearly with a lot of work put into it, and I would hate to see the article go through a WP:FARC. However, the article has multiple issues with references. Most notably, it's a 51 kb article with 5 inline citations and another 4 listed refs. That simply isn't enough references. Second, should those references be added (and the refs currently listed but not inline cited) they should really use inline citation to make it clear what is referenced from where and what is not. Finally, some sections such as the opening paragraph of "Original creation of the Laws" have clearly intended references (for sources I don't know, or I'd cite them) that should be converted to inline refs. I've informed Anville as the FAC nominator and listed maintainer, hopefully these issues get dealt with. Staxringold 11:48, 24 May 2006 (UTC)

I won't have time to deal with this until next week at the earliest, but hey, I was planning to re-write the Foundation Series article from scratch, so why not put some time in here too. Anville 15:38, 24 May 2006 (UTC)

OK, some of the problems were easier to fix than I'd expected. I'm out of time for today (and really, I did have more time-critical things to be working upon, things with looming deadlines like plumbous dirigibles). With the new footnoting scheme, further expansions and elaborations should be easier. Over the next few days, I'll get specific chapter and page numbers for the different items attributed to "Asimov (1979)" and "Gunn (1982)". I also have Joe Patrouch's book in my library now, which I didn't have when I first worked on this page, so a few new footnotes might well be appearing.

And many, many thanks to Raul654 for fixing the results of my brain failures. I promise not to make this particular mistake again, leaving only the infinite number of others I have yet to make. Anville 21:54, 24 May 2006 (UTC)

The article now has thirteen general references. Thirty-five footnotes direct the reader to specific pages of those references or to brief, stand-alone sources. Is there anything else I need to do? Anville 01:50, 1 June 2006 (UTC)

It looks great, thanks for the fixes! My only real remaining issue is the list section "Pastiches, parodies and adaptations", which can probably be split-off and just summarized here (removing a list and some of the article length). Staxringold _talk^contribs 17:59, 5 June 2006 (UTC)

I was thinking about doing that. . . give me a moment to think of a good summary text, and off I'll go. Anville 21:50, 5 June 2006 (UTC)

Well, that's done. Anville 22:00, 5 June 2006 (UTC)

First Law : Not in my Neighborhood!

A robot may not ... through inaction, allow a human being to come to harm.

Removing the double negative: A robot must interfere whenever a human being is being harmed.

Imagine having such a robot around you, interrupting you constantly: "Don't eat fat food - you'll get overweight! Don't drink coffee - you burn your taste buds! Don't go out - sunlight is harmful! Don't drive - it's dangerous! etc etc". And when your robot isn't around, you neighbor's will do the same to you (because the law says "a human being", not "the robot's owner").

Did anyone ever notice this catch ? — Preceding unsigned comment added by Whichone (talk • contribs)

This catch is the basis of Asimovs' novel The Naked Sun, and in a more general way underlies all the robot stories: humans become dependent on robots and are helpless without them. That's why Asimov's human societies that deliberately choose not to use robots survive and prosper, while the robot-using societies stagnate and die. In fact the robots themselves, as they become more sophisticated, decide that humans would be better off without them in the long run.

In Asimov's robot-using societies, there is no crime, because the robots wouldn't allow it. No one smokes or drinks or uses drugs, because the robots wouldn't allow it. There is a scene (in Robots and Empire) where two men visit a room where valuable things are stored. The room has no locks or any other crime-prevention devices, because robots do not allow crime. One of the men remarks to the other that if they happened to be carrying a blaster they could simply destroy any nearby robots and there would be nothing to prevent them from stealing the room's contents. The second man is disgusted that the first man could even think of such a thing and regards it as proof of his inferiority. Fumblebruschi 04:11, 5 July 2006 (UTC)

When they were programmed into a computer

In one of the books it says that the three laws were programmed into an actual computer with 'interesting' results - I think this deserves a mention--Therealchaffinch 15:47, 16 June 2006 (UTC)

Specifics, please. Perhaps you're thinking of the short story "The Evitable Conflict"? Anville 20:49, 20 June 2006 (UTC)

Flaw of the Third Law

The Third Law of Robotics states that "a robot must protect its own existence, as long as such protection does not conflict with the First or Second Law." But the Second Law says that "A robot must obey the orders given to it by human beings except where such orders would conflict with the First Law."

So let's see...if a robot's own existence is being threated by a human, the robot can't fight back, because it obeys the First Laws, above all the others, including it's right to protect himself. And if the human told the robot that it can't protect itself, the robot wouldn't be able to protect itself, because if it did, it would being disobeying the orders given to it by humans, thereby breaking the Second Law.

So basically, robots can't protect their own existence. — Preceding unsigned comment added by 63.254.152.87 (talk • contribs)

You assume that humans are the only things which can threaten a robot's existence. The Three Laws don't allow robots to defend themselves against a Frankenstein mob of anti-robot rioters (although the rioters would probably be too heated to give coherent orders and bring the Second Law into effect!), but robots could protect themselves against falling rocks, gamma rays and other non-human hazards perfectly well. Anville 16:02, 22 June 2006 (UTC)

Of course it could protect itself from natural threats. But what if the robot was given an order from a human to not try and protect itself from natural threats? Basically, what I'm saying is that a robot can't protect itself if a human gives it an order to not protect itself.

If you can think of a situation where a human would give such an order, you've got the plot for a story. That's one purpose for the Three Laws: to give a mechanism for inventing robot stories. Anville 16:28, 22 June 2006 (UTC)

It's called Runaround, and Asimov already wrote it. Two humans and a robot on (I beleive) Mercury. The robot, being the only one on the planet, and being important to the proper funtioning of the base, has a strengthed 3rd law. One of the humans casually tells it to gather some substance from the surface. The source of the substance is in a volvanic vent that has corrosive gasses in it. The strengthed 3rd law tells the robot to not get near the vent, while the 2nd law forces the robot to try to approach to aquire the substance. The laws equal out at a certain distance fromthe vent, and the roboot ends up walking in circles, unable to escape the logic loop. The humans eventually have to put their lives in danger to force a 1st law response (that overrides the other laws).

I've found that almost all his short stories are about ways 'around' the 3 laws. As mentioned elsewhere, the laws are dependant on the definition of terms.

What is a human? (A person who speaks with a certain accent)

What is 'harm' [done to a human]? (Physical, mental, emotional)

What if the situation requires a human to be harmed, how do you choose which?

etc

--12.110.196.19 04:03, 5 July 2006 (UTC)

Original creation of the Laws

In this section (at the beginning), is the repetition of the sentence part of the quote or is this vandalism?

" Before Asimov, the majority of "artificial intelligences" in fiction followed the Frankenstein pattern: "Robots were created and destroyed their creator; robots were created and destroyed their creator—". [1]"

Cheers, Lukas 00:51, 5 July 2006 (UTC)

Appearances in Pop Culture

There's a significant reference to the First Law in the final season (I believe) of Babylon 5. Where should that be mentioned? --Masamage 01:51, 5 July 2006 (UTC)

Shouldn't "Asimov's Laws" be mentioned in the lead

Throughout the text, the laws are referred to as "Asimov's Laws"; the page is also listed under Category:Eponymous laws. This strongly suggests that the lead should begin "The Three Laws of Robotics, also known as Asimov's Laws, ..." or similar. I presume that they are referred to as the "Three Laws" by Asimov throughout his fiction, and sometimes as "Asimov's Laws" during discussion of Asimov's work, but at any rate it wouldn't hurt if the usage could be clarified too. TheGrappler 02:49, 5 July 2006 (UTC)

Non-univesality of the Three Laws

These three laws aren't univerally applied in fiction.

I'm thinking in particular of the T1 robots from Terminator 3, and the conceptually identical "War Machines" from Doctor Who season 2 or 3, both of which seeomed to exist ONLY for the purpose of wiping out humanity. —The preceding unsigned comment was added by 202.12.233.21 (talk • contribs) 05:06, July 5, 2006.

The article in no way claims that the laws are universal outside of Asimov's fiction. There are way too many "Killer Robot" stories out there to justify such a claim. GeeJo ^(t)⁄_(c) • 14:52, 5 July 2006 (UTC)

I agree. Google it for notability and see what you come up with.