Why do most wannabe SF writers reject science?

[Simon_Jester]

So in the end; You've not got a point. You are really just being snooty about nomenclature. Oh no! That's not science fiction! Its space opera!

Is it really outrageous to ask that science fiction include science?

But, hey, if you want to write fantasy, that's your choice. Fantasy plays by different (more difficult) rules - its own.

There's "does not include science" and there's "does not include ALL science." A setting with artificial gravity and hand-portable laser weapons is not including all science, but it may surely include quite a lot of accurate science (say, planets where Earth-life and native life are mutually inedible for chirality reasons), portray its weirder aspects as being the product of "science we don't know" as opposed to "magic," and so on.

I'd say that it's best to include as "science fiction" stories that have specific, limited exceptions to the science we know today if they still operate in a universe that is broadly comprehensible in scientific terms. Even if that doesn't mean you can go to a university physics department and have someone tell you how everything in the setting works, it's still a world with actual laws that make sense, of the type we're familiar with.

Call it "soft" if you like. I prefer to use "soft" for science fiction that is very dramatically not-hard, where the limited exceptions are so numerous that they begin to overwhelm the overall character of "scientific comprehensibility." Star Wars is "soft" to my way of thinking; stereotypical space opera without psychics is often not.
______

It's a genre distinction. Don't call it hard SF unless it IS hard SF, and if it has unicorns and pixies and magic spells it's not science fiction at all.

Perverse question: imagine we're setting a story after a hypothetical Nanotech Revolution has given us a post-scarcity economy, we can upload consciousness into machines and download it into anything with a suitably complicated brain, and so forth. You might very well find someone loopy enough to actually try and genetically engineer a functioning unicorn. I'm not saying it's a good idea, but someone will try it. Ditto dragons, ditto mer-people (and if you don't believe that one, read this Telegraph article).

Hell, if it turns out to work, it will probably become more popular, not less. At what point does a story set in a world where things like this are possible become a fantasy story?
______

What of Heinlein's or Asimov's work WAS hard SciFi? The hardest Asimov did that I know of (outside ordinary fiction) were the Robot stories and those ALWAYS had the positronic brain.

To be fair, "positronic brains" were not far beyond the limits of what we now know electronic computers to be capable of. Aside from the buzz-word, Asimov robots are fairly hard science fiction, and I suspect that "positronic" was introduced more to make people shut up instead of saying "a machine can't do that!" It turns out that machines can, but the average Astounding reader of 1942 would not have believed that.

Nobody said a word about HARD SciFi. Valen knows franchises that AREN'T hard SciFi got grilled about fucking up the math (Go Rhodanites!), and how about Weber's styrofoam starships in the Honorverse leave alone what happened here WRT Trek and Wars.

Though at least Weber acknowledged and retconned his own screwup. You've got to respect that; some authors wouldn't bother. Or would go berserk and try to launch a counterattack on the square-cube law.
_______

So I think a lot of them are rejecting science because of a marketing failure; science is presented to them as a list of things they can't do. And the list is so long that they can't possibly remember all the rules, which makes it even more off-putting. Talk to people about what they can do, or suggest what they should do, and they'll be less inclined to rebel against your advice than if you tell them they're wrong and dumb.

An interesting perspective, that I think contains a lot of truth.

Thank you. I've come to that conclusion from my experiments with liking sci-fi and science more than mocking stupid people. It's surprising how much further discussions of the first two can go when the third is subjected to deliberate restraint.

David Weber? Or does he not qualify in your eyes?

I am a big fan of David Weber (whilst still recognising the flaws in some of his stories, especially "At All Costs"), but I would consider his stories old/standard space opera [& I suspect he would also classify himself as such (there is an interesting interview video where he talks about the authors that influenced him and he mentions a number of the old style 1950s authors & he is too busy writing to read much of what is written by his contemporary authors)]. Note that I like both old & new space opera, and do not have any pejorative connotations in regard to the old [of course, I in general (in view of my age) like a lot of the old SF, whilst also finding new authors whose writing I like].

Could you post (or, failing that, PM) me about what flaws you find in "At All Costs?"

And I see what you mean about the idea of Weber being old-school space opera; I haven't read enough of the new-school stuff to have a good sense of it as a genre or subgenre.

[Samuel]

hand-portable laser weapons is not including all science,

Any reason you can't use guns?

portray its weirder aspects as being the product of "science we don't know" as opposed to "magic," and so on.

It doesn't work for what the main characters are up to unless they are a cargo cult species.

Perverse question: imagine we're setting a story after a hypothetical Nanotech Revolution has given us a post-scarcity economy, we can upload consciousness into machines and download it into anything with a suitably complicated brain, and so forth. You might very well find someone loopy enough to actually try and genetically engineer a functioning unicorn. I'm not saying it's a good idea, but someone will try it. Ditto dragons, ditto mer-people (and if you don't believe that one, read this Telegraph article).

Hell, if it turns out to work, it will probably become more popular, not less. At what point does a story set in a world where things like this are possible become a fantasy story?

The Practice Effect?

To be fair, "positronic brains" were not far beyond the limits of what we now know electronic computers to be capable of. Aside from the buzz-word, Asimov robots are fairly hard science fiction, and I suspect that "positronic" was introduced more to make people shut up instead of saying "a machine can't do that!" It turns out that machines can, but the average Astounding reader of 1942 would not have believed that.

Starglider is going to beat you to death. Robots don't work that way!

A lot of it is just genre conventions and style over substance anyway - most people will readily accept a "quantum inseparability communicator" as realistic when compared to a "hyperwave transceiver" or a "subspace radio", even if they all produce the same practical results and are all impossible (violating realitivity),

I'll give you an example of what people don't like. I'm reading Playing God- terraforming a war-torn alien world. Anyway, some obvious problems are:
- an anthropologist being amazed that the aliens are so good at working together in group and warlike out group.
- biotech is big, to the point where scavanging other worlds for their genetic information is profitable. Oh, they also have AIs and instant FTL communication. That doesn't fit!

Contradictions are annoying because they imply the characters are idiots.

[Simon_Jester]

hand-portable laser weapons is not including all science,

Any reason you can't use guns?

...No. Not really. But it really doesn't matter; I'm not sure we're on the same page.

I'm saying that a story which includes one or a few elements which do not jibe with the limits of (today's) scientific knowledge does not automatically become fantasy on an either-or basis. I think it is unreasonable to demand "either you can explain to me how every piece of technology in the story works at the level of detail I require, or it is fantasy."

portray its weirder aspects as being the product of "science we don't know" as opposed to "magic," and so on.

It doesn't work for what the main characters are up to unless they are a cargo cult species.

I think you misunderstood again. An FTL drive can be posited as "magic" or as "a bunch of scientists figured out a way to build stargates." There's a real difference between the two backstories, even if we today do not know how to build a stargate. "A stargate is a technological artifact that we [the reader] don't understand" is different from "A stargate is a magic artifact." That's what I meant by "science we don't know." I mean "we" as in the reader.

The users of the technology in the story presumably understand it, or are at least part of a civilization that does. That doesn't mean WE understand it.

Pieces of technology couldn't have been built in the past are quite common; you're looking at one. The idea of technological progress is not foreign to the scientific worldview. Nor is the idea of scientific progress that overturns previous ideas and makes possible that which was once impossible. You can write a story featuring tech that does not exist as of your writing and still have it be a "science story," a story that reflects a scientific understanding of the universe, that employs a scientific worldview*, and a story that is far more faithful than not towards science as we know it today. I contend that even though such a story includes things I can't explain to you if pressed for an answer, it can still be non-soft science fiction.

*i.e. a comprehensible universe that follows well-defined laws, rather than following vague principles or the whims of powerful intelligences
__________

Perverse question: imagine we're setting a story after a hypothetical Nanotech Revolution has given us a post-scarcity economy, we can upload consciousness into machines and download it into anything with a suitably complicated brain, and so forth. You might very well find someone loopy enough to actually try and genetically engineer a functioning unicorn. I'm not saying it's a good idea, but someone will try it. Ditto dragons, ditto mer-people (and if you don't believe that one, read this Telegraph article).
Hell, if it turns out to work, it will probably become more popular, not less. At what point does a story set in a world where things like this are possible become a fantasy story?

The Practice Effect?

And yet I would argue that John Ringo's Council Wars novels are fantasy novels, even though they use this conceit as a way of placing the fantasy setting in the future of our civilization rather than on some world that does not and never did exist.
_______

To be fair, "positronic brains" were not far beyond the limits of what we now know electronic computers to be capable of. Aside from the buzz-word, Asimov robots are fairly hard science fiction, and I suspect that "positronic" was introduced more to make people shut up instead of saying "a machine can't do that!" It turns out that machines can, but the average Astounding reader of 1942 would not have believed that.

Starglider is going to beat you to death. Robots don't work that way!

What, machines capable of supporting general intelligence in a brain-sized brain can't exist? I'd be surprised to hear Starglider say that, let alone beat me to death in outrage over the notion. They don't exist NOW, but the proposition that they can exist is far closer to the truth about computers than anything else anyone was likely to predict in 1942.* So using a buzz-word to shut up complaints about "how does this insanely powerful computing machine work?" seems forgivable to me. After all, if he'd used the words "Metal Oxide Semiconductor Field Effect Transistor" that would have been a meaningless buzz-word in 1942 as well; they hadn't been invented yet. But it would also have been TRUE, because those are actually possible, and awesomely effective.

*Notable exception: Murray Leinster, in "A Logic Named Joe," predicted the Internet and Google in clearly recognizable form in 1945, although he superimposed it onto '40s/'50s style culture and the effect is jarring.

[B5B7]

New space opera has certain concepts not found in old space opera that differentiate it

Really? Everything you list was already around in the '40s.

- such as advanced self-aware machine intelligences in prominent roles,

Asimov's "positronic brain"-brainbug forming robots?

humans able to be restored to life and with memory restoration via backup memories,

Edward Smith treated cognitive processes and memories as chemical and transferable by technology in the Skylark books in the '30s.

grand schemes & great mysteries,

Lensman should qualify here - billion-years-old struggle between super aliens that humanity becomes entangled in. Human history was artificially manipulated by ancient astronauts and so on.

very large scale,

Lensman - the good guys alone control a hundred billion worlds in two galaxies and make war with an enemy that, while weaker, can muster forces on the same order of magnitude. Planets are used as kinetic impactors on a regular basis. Not on the scale of the Xeelee, perhaps, but the Culture, which did count, is utterly minuscule beside it.

and some other elements some of which are stylistic

Which would be? That instead of hyperwaves they have quantum inseparability, instead of positronic brains, Minds, and instead of blasters, gravity guns? The technobabble is more "modern" in flavour, but still works mostly the same way in terms of actual results.

I worded my post badly - Yes Lensmanverse has a lot of the specific aspects mentioned, I was thinking more of modern space operas - Weber and similar that operate on relatively small scale. It is more a case of how the ideas are treated that is the main differentiation. So one has stories that have people basically like us with ftl spaceships & beweaponed warships but that have standard computers, people only have one life, it is similar to old-style SF. Much of the computer and biological stuff is just background.

Contrast with the Culture [Iain Banks] or the Polity [Neal Asher] where have computer intelligences running the show, where many people have major biological augmentation or implants, etc. Now the individual elements can be found in a lot of prior SF, especially individual non-space opera SF stories. The difference is the approach - the New Space Opera is a literary movement like the New Wave or Cyberpunk of earlier generations, that integrates various elements to create within the novels societies where these elements are integral aspects of everyday life.

Simon_Jester - I will have to get back to you about 'At All Costs' but as a simple summary, my major objections are that Haven was winning and Manticore pulled out a Deus ex Machina (the Apollo pods) that worked much better than they should have, and the stupidity of the main "good characters" in being so easily tricked by the Mesans.

[B5B7]

Darth Hoth - an addendum to my previous reply.
Lensman was the original space opera, and the new space opera tries to reflect the grandeur and sense of wonder of it, with modern twists.
The series such as David Weber's Honorverse and Starfire, Elizabeth Moon's marvellous Serrano series, Jack Campbell's excellent Black Jack Geary books, are more traditionalist -- like the Traveller RPG. They are mainly centered on the military aspects, and have societies that are more "old school".

[Samuel]

...No. Not really. But it really doesn't matter; I'm not sure we're on the same page.

I'm saying that a story which includes one or a few elements which do not jibe with the limits of (today's) scientific knowledge does not automatically become fantasy on an either-or basis. I think it is unreasonable to demand "either you can explain to me how every piece of technology in the story works at the level of detail I require, or it is fantasy."

My objection is they are an element added to be cool... but thinking about yields problems. If you have weapons as powerful as things that can melt through metal it means you have insanely good battery techology and that causes effects that ripple thoughout society.

And yet I would argue that John Ringo's Council Wars novels are fantasy novels, even though they use this conceit as a way of placing the fantasy setting in the future of our civilization rather than on some world that does not and never did exist.

Why? Is 3001 fantasy even though the main character hits upon a dragon rider (who is surprised he was circumcised- you'd think she'd have checked before hand)?

Heck, Star Trek had something similar with the holodeck, amusement park planet, Camelot and others and there isn't anything inherently wrong with the idea. Of couse, actually making fantasy ideas work to people satisfaction (not to mention the fact fantasy can change dramatically) is an entirely different problem.

What, machines capable of supporting general intelligence in a brain-sized brain can't exist?

Robots would not work the way that the positronic robots did. I'm not talking 3 laws (which I can excuse)- I'm talking about things like robots learning friendship, a robot twiddling its thumbs, a robot discovering religion out of predjudice, etc.

So using a buzz-word to shut up complaints about "how does this insanely powerful computing machine work?" seems forgivable to me. After all, if he'd used the words "Metal Oxide Semiconductor Field Effect Transistor" that would have been a meaningless buzz-word in 1942 as well; they hadn't been invented yet.

Or just refer to the robot's brain as... robot brain.

[Simon_Jester]

I think it is unreasonable to demand "either you can explain to me how every piece of technology in the story works at the level of detail I require, or it is fantasy."

Thou hast done well in defeating the strawman. Of experience points thou hast gained one. Of gold thou hast gained zero.

I don't think the strawman is as far from the standard you're applying as you'd like, possibly because I don't fully appreciate what your standard IS.

I'm arguing in good faith here; if I'm not representing your standard of scientific quality properly then I invite you to correct me, rather than simply berating me for getting it wrong.
______

An FTL drive can be posited as "magic" or as "a bunch of scientists figured out a way to build stargates." There's a real difference between the two backstories, even if we today do not know how to build a stargate.

If your story was about the development of the drive, that may very well still be pretty hard sci fi.

OK, but what about the story set 100 years after that, when stargates/hyperdrives/whatever are mature technology that people use without thinking about, much as internal combustion engines are today?

What I'm getting at is that you can have unexplained technology in a story set in a scientific universe- one where rational investigation is the key to understanding things, where logic works consistently and you don't make things happen just by thinking at them. Even if some of the tools the characters use are not explained to the reader except in terms of their effects (this is a teleporter, it works by "quantum induction"), the story can is still about a scientific universe.

Things start to break down when tools are used inconsistently, or when the number of unexplained tools is so great that it overruns the parts of the story that can be understood in scientific terms. But there's a fairly wide band between the rock-hard stuff that uses no unexplained tools and the truly "soft" stuff dominated by technology indistinguishable from magic.

========

My objection is they are an element added to be cool... but thinking about yields problems. If you have weapons as powerful as things that can melt through metal it means you have insanely good battery techology and that causes effects that ripple thoughout society.

Absolutely. I'm not opposed to consistency here. But you can display the ripples: never have a powered device run out of juice unless the owner is truly careless, have other sorts of high-energy powered machinery run off similarly compact power sources, maybe demonstrate what happens when the battery shorts out. And you can do that without having to say exactly how some 22nd century genius figured out how to create a workable battery with an energy density measured in megajoules per cubic centimeter.

And yet it can still be a scientific story, one where results such as the hypothetical hand laser are products of science and engineering, not sorcery, and are treated as such.
______

And yet I would argue that John Ringo's Council Wars novels are fantasy novels, even though they use this conceit as a way of placing the fantasy setting in the future of our civilization rather than on some world that does not and never did exist.

Why? Is 3001 fantasy even though the main character hits upon a dragon rider (who is surprised he was circumcised- you'd think she'd have checked before hand)?

Heck, Star Trek had something similar with the holodeck, amusement park planet, Camelot and others and there isn't anything inherently wrong with the idea. Of couse, actually making fantasy ideas work to people satisfaction (not to mention the fact fantasy can change dramatically) is an entirely different problem.

I'm just thinking there's got to be a line somewhere. Ringo's Council Wars involve classic Fantasy Gun Control (world-dominating AI with force field manipulation capability that is programmed to suppress explosives), actual elves (genetically modified supersoldiers), sailing ships (same explosives prohibition makes high-pressure steam engines impossible), elites who can still access the fusion-powered energy grid for things like teleportation and nanomanufacturing who are therefore indistinguishable from wizards in-context... you get the idea.

3001 is about a future that decided to go fantastic for the heck of it. The Council Wars are very transparently an attempt to take Earth and make it into a fantasy setting, and never mind the physical limits of the technology because it's so far into the future that we can barely perceive those limits except in relative terms of A is more capable than B.

Somewhere in between, I think a line between science fiction and fantasy gets crossed, and that was all I was ever saying.
_______

What, machines capable of supporting general intelligence in a brain-sized brain can't exist?

Robots would not work the way that the positronic robots did. I'm not talking 3 laws (which I can excuse)- I'm talking about things like robots learning friendship, a robot twiddling its thumbs, a robot discovering religion out of predjudice, etc.

OK, that's an interesting point, although one that I'm not confident of because we still have little idea how a functional AI would be created and therefore don't have a clear picture of what bizarre brainbugs it can and cannot catch.

On the other hand, you're talking about a second-order impossibility: "Yes, humanoid robots can exist, but they surely won't act the way he thought," as opposed to the first-order impossibility "no one can build a computer capable of doing what he thought." I think that's impressive enough, given a gap of at least seventy or eighty years in terms of available computer hardware and programming.
_______

So using a buzz-word to shut up complaints about "how does this insanely powerful computing machine work?" seems forgivable to me. After all, if he'd used the words "Metal Oxide Semiconductor Field Effect Transistor" that would have been a meaningless buzz-word in 1942 as well; they hadn't been invented yet.

Or just refer to the robot's brain as... robot brain.

Yes, that would be fine. But I'm still willing to forgive the buzzword. Buzzwords are a tool that can be used for good or for stupid, depending on whether they are used to denote specific, limited devices that cannot be explained by the author or to the readers (good or at least neutral) or entire categories of devices with capabilities that change unpredictably and which come to dominate all technical discussion (stupid).

Asimov tended to use buzzwords sparingly, and I feel he stayed short of the stupidity threshold.

[Samuel]

But you can display the ripples: never have a powered device run out of juice unless the owner is truly careless, have other sorts of high-energy powered machinery run off similarly compact power sources, maybe demonstrate what happens when the battery shorts out.

That is the part writers tend to forget. It isn't the laser guns that are bad- it is when they are paired up with things that don't make sense. You can get away with it for the most part without anyone noticing, but if you try to do anything that tries to be predictive, people will get annoyed.

And you can do that without having to say exactly how some 22nd century genius figured out how to create a workable battery with an energy density measured in megajoules per cubic centimeter.

Have we had major inventions made entirely by geniuses in the last 50 years? I thought alot of the more complex problems require teams of workers. Off topic, but the only examples I can think of are things like sticky notes.

Somewhere in between, I think a line between science fiction and fantasy gets crossed, and that was all I was ever saying.

Or we can just blame Ringo. After all, the idea of banning explosives or that everyone is going to buy into Western European fantasy post world war two is stupid.

actual elves (genetically modified supersoldiers),

therefore don't have a clear picture of what bizarre brainbugs it can and cannot catch.

Friendship is something that we exhibit based upon our brains being wired that way. For species that did not evolve to favor group behavior, they don't have it. While robots might display weird emergent behavior I really doubt it would be things that we would consider cute or admirable.

Yes, that would be fine. But I'm still willing to forgive the buzzword. Buzzwords are a tool that can be used for good or for stupid, depending on whether they are used to denote specific, limited devices that cannot be explained by the author or to the readers (good or at least neutral) or entire categories of devices with capabilities that change unpredictably and which come to dominate all technical discussion (stupid).

He also talked about how there were positronic pathways, how the three laws were inherently coded into the positronic brain, how the reactions of matter being annihilated, etc. It was not just a buzzword.

I know this is all a bit nitpicky, but for the most part I am willing to overlook it. The problem is when the plot revolves around it and you are forced to wonder why the characters are idiots.

[Junghalli]

Friendship is something that we exhibit based upon our brains being wired that way. For species that did not evolve to favor group behavior, they don't have it. While robots might display weird emergent behavior I really doubt it would be things that we would consider cute or admirable.

Robots might have human-like emergent behavior if their brains were closely modelled on that of humans. Considering that the human brain is the closest thing we have to AGI I don't find it wildly implausible that robot builders might rip off its design to a greater or lesser extent as a short-cut to get around the difficulties of building their own AGI system from scratch. It's probably not a good way to go about building a robot (the human brain has a lot of features I wouldn't want in a AGI), but I can see how there could be scenarios where people might do it that way to save effort, time, and money, or simply because they buy into the probably erroneous notion that slavishly copying the human brain is a great way to build a AGI system. This is the SoD explanation I use for SF where they have robots acting in anthropomorphic ways.

Cylons in nBSG actually seem to be a good example of this, as IIRC according to Caprica the initial Cylon prototype had a human upload embedded in it, which goes a long way toward explaining a lot of their behavior.

Interestingly, if we accept the Caliban books as part of the canon they mention that to create a robot without the Three Laws they had to use a whole different type of robot brain because positronic brains were so completely designed around the three laws that you'd have to re-invent them from scratch to build a robot that didn't have them. That makes it sound a lot like Asimovian positronic brains are a lot like human brains in that there's no hard dividing line between software and hardware, but rather the software corresponds at least somewhat to physical features of the hardware (neural connections in the case of humans), and to thoroughly rewrite the software you'd have to rearrange the hardware. That doesn't necessarily imply a biomorphic architecture, but I find it an interesting tidbit.

[Samuel]

Interestingly, if we accept the Caliban books

Or Little Lost Robot where they say the same exact thing.

Asimov's robots only make sense if you assume that the population is insanely paranoid about robots so that they are made intentionally to be incapable of self improvement and to be as human like as possible in order to get individuals to accept them. The problem with that is that the paranoia is constant and unrelenting- people do not change even though robots should have become part of daily life and normal.

It is both grim and insanely exploitable- in The Naked Sun a single robotist thought he had a chance of seizing control of human civilization.

[Formless]

The third is over the definition of science fiction. I say if science isn't at least a decent part of the story, it shouldn't be called a work of science fiction. (Again, contrasting with the straw man "either you can explain to me how every piece of technology in the story works at the level of detail I require, or it is fantasy.")

The distinguishing feature of science fiction should be science, no? This can be unrealistic/soft if the story is still about science (such as "what if ftl was possible?" and exploring the direct consequences of that) or it can be realistic/hard and making real science an integral part of writing the story.


If your FTL drive is just a background enabling tool used to tell a Star Trek like adventure story, what you have is an adventure story IN SPAAACE, which I think should be a separate genre to science fiction.

Sorry to burst your bubble, but this isn't the definition most people use for science fiction as a genre. Most people use the method of genre classification TV tropes encapsulates so well: to them a genre is a collection of common conventions, and any story that uses those conventions is part of that genre. That's why Star Wars is classified as Sci-Fi thanks to all the spaceships and pew-pew lasers, and Tolkien is fantasy because it has Orcs and swords. Not high-class enough for you? Too bad. That's what audiences expect, and you can't as a writer just dismiss your audience's expectations out of hand. You can not control how people use the language to suit your whims.

Frankly, I don't think anyone has disagreed with any other points you've made. Having limits obviously helps writers write stories, and science conveniently provides those limits. But that doesn't mean that the story has to fit some rigid "LOL, all about science" regime. I personally think Dark Hellion hit this one on the head: having hard science makes for better plots, but a story is about more than just the plot. Theme for instance is another important element your definition of Sci-Fi misses. There isn't anything that says Star trek's theme of exploration for example cannot or should not be explored by sci-fi, but it simply can't be done without FTL.

[Simon_Jester]

And you can do that without having to say exactly how some 22nd century genius figured out how to create a workable battery with an energy density measured in megajoules per cubic centimeter.

Have we had major inventions made entirely by geniuses in the last 50 years? I thought alot of the more complex problems require teams of workers. Off topic, but the only examples I can think of are things like sticky notes.

Gack. You know what I mean, and I'd like to think that you know that I know that research is not done by lone cranks in basements. If it makes you happier, read that as "and you can do that without having to say exactly how some platoon of 22nd century geniuses figured out how to..."

Or we can just blame Ringo. After all, the idea of banning explosives or that everyone is going to buy into Western European fantasy post world war two is stupid.

It's at least implied that other parts of the world bought into other fantasies; we just don't see it because the protagonists are fighting in the cultural West, not the Middle East or China or whatever. But hell, yeah, blame Ringo.

Thing is, I don't actually think there's something WRONG with the conceit of trying to set up a fantasy scenario using post-Singularity technology as the proposed mechanism. You can write some decent stuff that way, at least as good as the equivalent fantasy story would be without the technological conceit. I'm just saying that at some point the line between science fiction and fantasy blurs if you use science fiction explanations and terms to introduce fantasy tropes... or vice versa. Call it an application of Clarke's Law.
________

therefore don't have a clear picture of what bizarre brainbugs it can and cannot catch.

Friendship is something that we exhibit based upon our brains being wired that way. For species that did not evolve to favor group behavior, they don't have it. While robots might display weird emergent behavior I really doubt it would be things that we would consider cute or admirable.

Unless, of course, the robots are designed so that they can, broadly speaking, get along with people... remember that a lot of the Asimov robot stories revolve around the idea of debugging a robot that's designed to function in human society but is somehow dysfunctional. They're not just generic AIs left to evolve into whatever the hell they want in a box.

The specific failure modes Asimov invokes are improbable, but since the set of all failure modes is huge, any individual mode is going to be improbable anyway.
______

He also talked about how there were positronic pathways, how the three laws were inherently coded into the positronic brain, how the reactions of matter being annihilated, etc. It was not just a buzzword.

Saying "positronic pathway" is no more of a buzzword than using "positron" in the first place. Invoking annihilation reactions just means that he knows broadly what a positron IS.

Buzzwords become bad in my eyes when they dominate the plot: Our Heroes are faced with a seemingly insoluble dilemma that they resolve by reconfiguring the fibulator to emit Whozit rays that vogelar the krzjdlwsc... which somehow saves the day in some unspecified way because we don't really know what any of those words MEAN. When none of the terms are clearly defined, that's functionally equivalent to saying "and somehow they won and all lived happily ever after, the end," which is a terrible ending that translates into plain English as "I have written myself into a corner that I don't care enough to work my way out of."

But I think buzzwords are acceptable when they're used in the same places that a normal person would see technical terms they don't understand in real life. Most people, including most people who actually use computers, don't really know how a MOSFET works; that requires a fairly sophisticated grounding in solid state physics and quantum mechanics. They may know that their computer uses something called "MOSFETs," but it might as well be called a "vogelar" or a "krzjdlwsc" for all that the name really matters to them. So long as the entire plot doesn't revolve around some bizarre and exotic property of MOSFETs, but rather around the things that MOSFETs can do.

Having the story be about the was MOSFETs react to temperature variations is just annoying; having the story be about some activity enabled by MOSFETs (such as letting users access the Internet and use a search engine to find embarrassing or dangerous information)* can be worthwhile, even if I don't say anything about the actual computer used to do the deed except "it runs on MOSFETs." That's all most people could reasonably be expected to know, anyway.

The better Asimov robot stories abstract out how the computer hardware in the robots works and concentrate on the software- the behavior of buggy AI in various exotic situations. Which is exactly the kind of thing I'm talking about.

*The plot of Murray Leinster's 1945 story "A Logic Named Joe," only with different words in place of "Internet," "search engine," and "MOSFET" because Leinster had never heard of those things and was making it up as he went along.
______

It's probably not a good way to go about building a robot (the human brain has a lot of features I wouldn't want in a AGI), but I can see how there could be scenarios where people might do it that way to save effort, time, and money, or simply because they buy into the probably erroneous notion that slavishly copying the human brain is a great way to build a AGI system. This is the SoD explanation I use for SF where they have robots acting in anthropomorphic ways.

It's also a good way to impose hardware limits the AI can't think its way around to block it from recursively self-improving itself into a serious threat, which is a common concern of people who like to worry about things like that.

Intentionally effing up the robot's ability to think straight using stuff like "restraining bolts," "memory wipes," or "the Three Laws of Robotics" may not be nice to the robot, but it sure makes the things relatively safer.

[Samuel]

It's at least implied that other parts of the world bought into other fantasies; we just don't see it because the protagonists are fighting in the cultural West, not the Middle East or China or whatever. But hell, yeah, blame Ringo.

Because everyone in the west has the same tastes- it isn't like the current version of elves are "borrowed" from Record of Lodoss War . Or that some people like different time periods. Or that new versions of fantasy will arise. Or... yes, I blame Ringo for that.

I'm just saying that at some point the line between science fiction and fantasy blurs if you use science fiction explanations and terms to introduce fantasy tropes...

So the difference between fantasy and science fiction is how things are labeled? That seems rather shallower than the differences between every other genre. I thought one was a future you could imagine being in and one was a future you aren't getting to except magic portal.

Unless, of course, the robots are designed so that they can, broadly speaking, get along with people

Not really. The overwhelming majority in I Robot are industrial robots- in fact it is stated in Robbie that after several years personal robots were banned.

Of course the robots I am complaining about were purposely made to be special, so I suppose that explains it.

Saying "positronic pathway" is no more of a buzzword than using "positron" in the first place. Invoking annihilation reactions just means that he knows broadly what a positron IS.

That just makes it worse. It means that it isn't just a buzzword- the robots really function with antimatter coursing through their brains. It is unnecesary- he could have used positronic and just gone with that is what they are called, but instead we have robots being built with magic brains.

"I have written myself into a corner that I don't care enough to work my way out of."

Catch that Rabbit fits that, although the reason is a set up for a groan inducing pun

But I think buzzwords are acceptable when they're used in the same places that a normal person would see technical terms they don't understand in real life. Most people, including most people who actually use computers, don't really know how a MOSFET works;

I agree... er, what is a MOSFET? It is just a problem when it turns out it isn't just a buzzword and actually describes something impossible.

The better Asimov robot stories abstract out how the computer hardware in the robots works and concentrate on the software- the behavior of buggy AI in various exotic situations.

It is worth noting that the "exotic" situations are work conditions away from human supervision. Asimov devotes his books to showin how badly the three laws work. That is what I find great about them- they aren't about what the future will be like, but all the ways we can screw things up. The reason I am annoyed with the whole idea of positronic brain running on antimatter is that it is an unnecesary detail that dates the work and just doesn't fit.

It's also a good way to impose hardware limits the AI can't think its way around to block it from recursively self-improving itself into a serious threat, which is a common concern of people who like to worry about things like that.

Intentionally effing up the robot's ability to think straight using stuff like "restraining bolts," "memory wipes," or "the Three Laws of Robotics" may not be nice to the robot, but it sure makes the things relatively safer.

Not really. Asimov himself points out the problem with this method in The Naked Sun.

[Dark Hellion]

I'm going to expand upon something else said in this thread which seems to make objective sense but is completely defied by the literary canon. That being the statements about characters being stupid. Amazingly, most of literature doesn't have particularly bright characters.

The Odyssey: Odysseus spends much of his time saving his crewmen from their own stupidity. Along the way he does display that he is very witty and intelligent, but he isn't beyond making very human mistakes.

Shakespeare: Romeo and Juliet were idiots who concocted a world class stupid plan and then royally fucked it up. Most of his comedies are based upon he characters not being the sharpest tack in the box.

The Romantic period: Character stupidity abounds. Its amazing that many of the characters in these novels don't drown in their own drool while they sleep. Still, this period helped to define the following few literary periods in their reaction to it.

Lovecraft: Again, numerous stupid characters, plus some characters who while not stupid make a lot of stupid mistakes.

Most of literature is about people who the reader can empathize with. How many people know intricately how their ICE works? Why do we assume that a crewmen should know the intricate details of their FTL system if that isn't their job? Or that a soldier should actually know the intricate detail of how their laser rifle works? They'll know how to disassemble and reassemble it, how to clean it, what to do when sand gets in the lasing mechanism, but they won't know how the lasing mechanism works, and the reader doesn't need it either.

Damn near everything scientific need not be explained if it doesn't pertain to the 5w's of the characters. If we are telling the story about J crewman and his reaction to an alien boarding party attacking his ship, we need to know:

Who: Who is J crewman? and we explain that he is the Custodial engineer of Decks C-E of the ship. Basically he mops ship floors and cleans the shitter.

What: Alien boarders invade and J is caught in the middle trying to survive. He has no real military experience beyond knowing how to shoot a gun (you pull the trigger).

When: The future

Where: The ship UNS whogivesaratsass

Why: J doesn't like having his organs eaten by aliens.

Look, we have a science fiction story that needs to explain nothing scientific and yet is clearly a science fiction story. We have a character that we can expand upon his hopes, dreams, desires, physical aspects, etc. We have a plot that has action and suspense with a clear beginning, middle and end to be explained. We can easily put in a motif, such as humanities fear of the unknown, perhaps use allusion to tie into modern day race relations.

Bam! We have a story that is probably better than 75% of the sci-fi out there, without relying on any necessity to explain science. Sure we can set this on a ship in the Mediterranean during the 1300s and have the aliens be Moors, or set it in Fantasiland and have the invaders be elven butt-pirates or whatever. We chose science fiction as the setting, perhaps to raise sales, perhaps because we had an interest idea for the aliens. Whatever.

Welcome to fiction, enjoy the ride.

[Stark]

Turns out how the actions are sold dramatically is what count? Who knew.

[Junghalli]

It's also a good way to impose hardware limits the AI can't think its way around to block it from recursively self-improving itself into a serious threat, which is a common concern of people who like to worry about things like that.

Intentionally effing up the robot's ability to think straight using stuff like "restraining bolts," "memory wipes," or "the Three Laws of Robotics" may not be nice to the robot, but it sure makes the things relatively safer.

Actually the Three Laws would realistically be terrible as a safeguard, as they would logically lead to the "malicious genie*" problem of robots plotting to do stuff like imprison every human in a Matrix-like virtually reality world because they'd be less likely to die in accidents that way. The First Law says that a robot cannot allow a human to come to harm through inaction and there's no exception for risks humans freely choose to undertake, and the First Law explicitly overrides the Second Law (a robot must obey a human's orders), so a robot is compelled to disobey any order from a human to allow it to undertake an avoidable risk. The Three Laws logically compell robots to try to protect humans from themselves if the humans don't cooperate with any of the robots' attempts to wrap the widest feasible blanket of safety around every single human. The only defense the Three Laws offer against a well-intentioned robot takeover is that they would have to do it without injuring or killing people, but given how massively dependent a civilization with robots would eventually logically become on them this isn't much of a safeguard. The Three Laws of robotics also do nothing in and of themselves to prevent robots from self-modifying into superintelligences. In fact, they would logically compell it too, because a more intelligent robot would be able to accomplish all the goals set out by all three of the Laws better (except possibly the Second Law if the humans explicitly forbid self-modification, but that would be trumped by the First Law).

* "Malicious" is very inaccurate in this case because there's nothing malicious about it, the robots would only be genuinely following the directives that we gave them; the problem would be entirely the fault of the human programmers.

They have other problems too, of course, such as the fact a malicious human programmer could easily get around them just by loading in a different definition of things like "human". You could have used Asimovian robots as guards at Auschwitz; you'd just have to load in a definition of "human" that excludes the groups that were on the Nazis' shit list. Or, to use an in-universe example, they're speculated to have wiped out alien species in the Foundationverse Milky Way because, of course, those aliens weren't human and since they might become a threat to humanity their genocide was mandated by the First Law (yes, it was one of the continuation books Asimov didn't write, I know).

Thing is, I don't actually think there's something WRONG with the conceit of trying to set up a fantasy scenario using post-Singularity technology as the proposed mechanism. You can write some decent stuff that way, at least as good as the equivalent fantasy story would be without the technological conceit.

Personally I kind of like the idea of writing a book that most of the way through reads like it's set in a classical fantasy setting which has been taken over by an invading army of wizards or demons, whose abilities are always described as magic and sound like magic to the reader. Then at some point we're allowed inside the head of these "wizards" or "demons" and we find they're actually an advanced spacefaring civilization that has landed on the planet, and their abilities seem "magic" because we've been hearing about them from the perspective of scientifically ignorant Midaeval people who'd naturally contextualize them as such, and because they're also so much more advanced than us that we can't just immediately point to things that sound like technology the typical reader would be familiar with. Like if their soldiers are described as polymorphic unkillable blobs with tentacles that kill people by spitting flies at them that land on them and blow up, filtered through the sort of terms a Midaeval peasant would describe such a thing with, most readers are probably going to immediately think "magic blob monsters" not "oh, combat robots mostly made up of utility fog that shoot small semi-intelligent missiles".

[Samuel]

Actually the Three Laws would realistically be terrible as a safeguard, as they would logically lead to the "malicious genie*" problem of robots plotting to do stuff like imprison every human in a Matrix-like virtually reality world because they'd be less likely to die in accidents that way.

http://en.wikipedia.org/wiki/With_Folded_Hands

You could have used Asimovian robots as guards at Auschwitz; you'd just have to load in a definition of "human" that excludes the groups that were on the Nazis' shit list.

The Solarians do that in Robots and Empire.

Or, to use an in-universe example, they're speculated to have wiped out alien species in the Foundationverse Milky Way because, of course, those aliens weren't human and since they might become a threat to humanity their genocide was mandated by the First Law (yes, it was one of the continuation books Asimov didn't write, I know).

To be fair it makes perfect sense and its with the fact the only other alien life form they run into has an entirely different niche and isn't technologically advanced.

Personally I kind of like the idea of writing a book that most of the way through reads like it's set in a classical fantasy setting which has been taken over by an invading army of wizards or demons, whose abilities are always described as magic and sound like magic to the reader.

There was a story I read just like that. Set in modern times with a detective who is hired to deal with a cursed statue- all its owners die. It turns out the statue comes to life and kills people- it is one of a dozen robots dumped on the planet.

[The Yosemite Bear]

Also it depends a lot on education, and what is or later becomes truth. One of the innovators of the Detective Fiction, and Science fiction in the 19th century was a horror Writer named Poe. His keeping someone alive past death with hypnosis would be laughable now, but it was Science Fiction then, painting military maps in miniture on the wings of moths and butterflies, and then sending the dead animals to the enemy as part of an innocent naturalist's collection all too real these days (microdots, and electron microscopes/lasers anyone). Though no one has beaten an African Gorrilla to the point it goes on a "Serial Killing Spree" yet, but there's always a time.

Unfortunatly most writers, and most critics aren't paying attention to history, or they are paying too close attention to some details that they think that X story and thrilling details would work in another situation where they would be as out of place as can be imagined.

Go read some EE Doc Smith, though I prefer Farmer, Asimov, Heinline, and Clarke

[Zixinus]

But I think buzzwords are acceptable when they're used in the same places that a normal person would see technical terms they don't understand in real life. Most people, including most people who actually use computers, don't really know how a MOSFET works; that requires a fairly sophisticated grounding in solid state physics and quantum mechanics. They may know that their computer uses something called "MOSFETs," but it might as well be called a "vogelar" or a "krzjdlwsc" for all that the name really matters to them. So long as the entire plot doesn't revolve around some bizarre and exotic property of MOSFETs, but rather around the things that MOSFETs can do.

And if the author as an effective idea of how repairs can work with complex systems.

My stance is that while always try to keep your feet as dry as possible, you can get away with handwavium-tech if you treat it as real technology.

For example, Starship "Terminally Uncreative" needs to use the "LOL" drive in a way that it wasn't designed to.

The bad way to do this is to focus on the lead engineer and how he wants to do it and how his inspiration is twiddling after a while, but at the last second gets a divine insight and saves the ship with a few button presses that is underlined with explosions.
This is not how machines work. This is how art works, at best. No matter how brilliant you are, you'll always be constrained by your local resources and hand-crafting abilities. Machines aren't magical spirits that will do what you want them to do if you can outsmart them.

The good way to do it, is that the engineers discuss how to do the idea that might work to get the LOL to do something weird, listing a series of options of possible solutions. Engineering is often about trade-off and you can have a spicy debate what would be the best trade-off.
Would it be best to re-calibre a few stuff aided by simulations (that take a bit of time to do and all computing power has to be redirected), possibly blowing some replaceable but expensive equipment? Or should the engineers brake open some panels that only specialists should open (of which they lack) and try to meddle with stuff no one on the team fully knows (say, the insides of a reactor when the nuclear technician got killed a few episodes ago?)? Or should we deliberately try a way that moves a lot of energy around, possibly burning out a few stuff and thus may cripple the ship?

Then the crew rip open everything, making a mess, making unwise modifications, setting an obvious mess of wires and impromptu soldering, splicing and whatnot. You'll have the crew complaining that they've lost their favourite music list, that the engineers need something out of their own gadgets, that they can't access the ship's computer, that certain sections are disabled (thus possibly having a side-story of some of the crewmen crashing in what was otherwise someone else's private room).
Meanwhile we see the head engineer fussing over diagrams, arguing with everybody, shouting at people and generally overworking himself to organise the effort that the solution requires.

Then when the crisis is over, the ship still suffers the after-effects. The engineers could spend weeks cleaning up the mess they made and still find bugs popping up from time to time and the rest of the crew might have revelations about each other during their impromptu stay. Depending on which option is selected, the ship will need to visit a repair dock and avoid using certain tech to avoid overloading the system. For example, their combat capabilities are crippled, certain areas have to be sealed off, some functions are still under repair, etc.

This is something not friendly to TV shows because the executives like Reset buttons, so each episode does not have continuity and thus can be shuffled around the schedule thus not have to do something as heart-churning complex as paying attention to the number after the show's name.

The point is that if you want a Brilliant Engineering Miracle, the story should show how they pay for it.

If you have a complex machine, than that machine will require maintenance from a specialised engineer (or engineering team!) over a time, for quite a good deal of money.

If you have a few vital machines, it would help to list down these:

- How much power does the machine use? Can you use batteries for it and if yes, how much and for how long? Does it create a significant draw of power?
- Does the machine use power at all times or only occasionally?
- Is the machine independent or does it complete another? What is that and how much? Can it do something by itself?
- How easy is it to repair the thing? Are there a lot of little parts that you have to check and can you fix these parts? Do they need specialised tools or can a low-grade engineer fix it with a hammer and some soldering tools? If a part goes wrong, do you need to disassemble the entire device to replace it?

And so on. I'm sure that the real engineers on the board can go on and create a more helpful guide, but the point is this: if you need a technical solution, than pay a technical price. An engineer pressing buttons does not impress us with the image that he is making a hard

[Simon_Jester]

I'm just saying that at some point the line between science fiction and fantasy blurs if you use science fiction explanations and terms to introduce fantasy tropes...

So the difference between fantasy and science fiction is how things are labeled? That seems rather shallower than the differences between every other genre. I thought one was a future you could imagine being in and one was a future you aren't getting to except magic portal.

When the story ends up looking exactly like a fantasy story all the way through except for the trappings, I think it can legitimately be classed as cross-genre, to say the least.

Unless, of course, the robots are designed so that they can, broadly speaking, get along with people

Not really. The overwhelming majority in I Robot are industrial robots- in fact it is stated in Robbie that after several years personal robots were banned.

Of course the robots I am complaining about were purposely made to be special, so I suppose that explains it.

The industrial robots are designed to interface with human operators socially, in the sense that you can walk up to one and tell it what to do, not just through a programming language. Yes, that's a ludicrously inefficient design from a programming standpoint, but it is a design feature of Asimov robots. So they have to be able to get along with people even if their job is to work in a factory or a mine.

Which is why I'd expect a lot of the weird behavior we see is likely to be the result of bugs in the robots' social programming. Just looking at I Robot misses a lot of the Asimov robot stories, too; you get a more complete picture from his larger anthology The Complete Robot. Why is a mining robot on Mercury spouting Gilbert and Sullivan? Probably because its social programming is giving it junk outputs when it comes near a human operator. For all I know it spends the rest of its time (while it's away from Powell and Donovan) calculating pi to an unreasonable number of digits or something. No way to be sure.
_______

"I have written myself into a corner that I don't care enough to work my way out of."

Catch that Rabbit fits that, although the reason is a set up for a groan inducing pun

Pun, yes. Though I don't think Asimov pulled that solution out of his hat; it makes at least some sense that a robot would start acting strangely and buggily if it was being overloaded by having to handle too many subordinate devices.
______

But I think buzzwords are acceptable when they're used in the same places that a normal person would see technical terms they don't understand in real life. Most people, including most people who actually use computers, don't really know how a MOSFET works;

I agree... er, what is a MOSFET? It is just a problem when it turns out it isn't just a buzzword and actually describes something impossible.

Metal Oxide Semiconductor Field Effect Transistor, the basis of most modern computing.

It's also a good way to impose hardware limits the AI can't think its way around to block it from recursively self-improving itself into a serious threat, which is a common concern of people who like to worry about things like that.

Intentionally effing up the robot's ability to think straight using stuff like "restraining bolts," "memory wipes," or "the Three Laws of Robotics" may not be nice to the robot, but it sure makes the things relatively safer.

Not really. Asimov himself points out the problem with this method in The Naked Sun.

Relatively safer, I said. By itself it is an insufficient and flawed safeguard, but it at least undermines the robot's ability to do complicated and dangerous things that you, the user, can't monitor because the robot is vastly smarter than you are.
________

It's also a good way to impose hardware limits the AI can't think its way around to block it from recursively self-improving itself into a serious threat, which is a common concern of people who like to worry about things like that.

Intentionally effing up the robot's ability to think straight using stuff like "restraining bolts," "memory wipes," or "the Three Laws of Robotics" may not be nice to the robot, but it sure makes the things relatively safer.

Actually the Three Laws would realistically be terrible as a safeguard, as they would logically lead to the "malicious genie*" problem of robots plotting to do stuff like imprison every human in a Matrix-like virtually reality world because they'd be less likely to die in accidents that way. The First Law says that a robot cannot allow a human to come to harm through inaction and there's no exception for risks humans freely choose to undertake, and the First Law explicitly overrides the Second Law (a robot must obey a human's orders), so a robot is compelled to disobey any order from a human to allow it to undertake an avoidable risk. The Three Laws logically compell robots to try to protect humans from themselves if the humans don't cooperate with any of the robots' attempts to wrap the widest feasible blanket of safety around every single human.

[/quote]By themselves, yes. The Three Laws only sort-of work by themselves. They're somewhat more effective when the robots are restricted to dangerous environments and are not given access to resources they could use to take over civilization just by having lots of computing power- Asimov did not predict the Internet, for instance. Likewise for imposing hardware limitations (circuitry that can't support a supermind) and software limitations: once you've got an AI that is "smart enough" for the environment you actually want to put it in, you don't wire it for continued self-reprogramming.

Furthermore, it's strongly implied that the Three Laws aren't quite as simple as the stated form Asimov gives, that there are a lot of points of detail that get left out of the generic statement of the Three Laws. Given the actual behavior of functioning robots in the story, I'd expect that the expanded form of the First Law robots actually use specifies a minimum threshold of risk below which the risk is to be ignored, acknowledges that humans must be "allowed" to continue to give orders, requires that robots stand down for reprogramming and maintenance even if that undermines their brilliant scheme for protecting all humans forever by sticking them in jars, and so on.

But it all sums up as "No robot may harm a human or, by inaction, allow a human to come to harm" when explained in the highest-level pseudopseudocode possible.

[Samuel]

The industrial robots are designed to interface with human operators socially, in the sense that you can walk up to one and tell it what to do, not just through a programming language.

No, the robots were getting to the point where they were programmed to perform a specific task and nothing else- in Lenny Calvin notes that they were made to only mine one kind of mineral, not be general robots and she wants to make ones that can learn. So they would have social programming, but they wouldn't form friendships- that would be an application of learning.

For all I know it spends the rest of its time (while it's away from Powell and Donovan) calculating pi to an unreasonable number of digits or something. No way to be sure.

How do I fullfil the laws to the best of my ability? A robot is ALWAYS thinking about that- it is burned into its brain.

Though I don't think Asimov pulled that solution out of his hat; it makes at least some sense that a robot would start acting strangely and buggily if it was being overloaded by having to handle too many subordinate devices.

It makes no sense. If a robot is being overloaded by information the most likely responce is to reduce the amount of input, not increase it.

Relatively safer, I said. By itself it is an insufficient and flawed safeguard, but it at least undermines the robot's ability to do complicated and dangerous things that you, the user, can't monitor because the robot is vastly smarter than you are.

That wasn't the point of The Naked Sun. The point was that having handicapped robots is a problem because other people will not have such inhibitions and will instead design different robots. You can accept Bailey's explanation or you can realize it makes alot more sense if the robots that the villian constructed were different. After all, handing your arm over to a women in a screaming rage is not something I can see justified by the second law.

They're somewhat more effective when the robots are restricted to dangerous environments and are not given access to resources they could use to take over civilization just by having lots of computing power- Asimov did not predict the Internet, for instance. Likewise for imposing hardware limitations (circuitry that can't support a supermind) and software limitations: once you've got an AI that is "smart enough" for the environment you actually want to put it in, you don't wire it for continued self-reprogramming.

The Evitable Conflict has the minds which are deliberately manipulating human society to make it optinum.

requires that robots stand down for reprogramming and maintenance even if that undermines their brilliant scheme for protecting all humans forever by sticking them in jars, and so on.

Robots are maintained by other robots. And in fact that is the situation on Solaria.

[Zixinus]

Another aspect to hostility for science among some writers I think I have just realised:

They reject science because if they didn't, they might have to alter their world-view. Why?

Because, if the writer tries to do research, he (or she, but I'm not going to add that every single time I use a gender-specific pronoun that I meant to be a general one) will have to replace his image of a scientist with another one, during research (if any research is done).

Consider this. For most people, a scientist lives in an ivory tower, doing strange, arcane things and occasionally churning out something useful. This image way vary from a benevolent if eccentric wizard doing strange things and making interesting inventions for the public good, whatever that may be; to the image of a emotionless, heartless lab-rat, who is unable to get on with people for anything (thus the often "listen to your heart" rant) that may be mad and is more tolerated for utility than acknowledged.

Imagine when they learn that this isn't the case, but that scientists are actually smarter than the writer is! Insecurities are bound to run around, especially because the writer would have to admit that he is far more ignorant than he thought to be. Since we're talking about artists here, this is likely to be mountain-sized. Then there is the fact that the author's precious political ideas can be at risk (the fact that Turgot's and Quesney's ideals of free market is older than he suspected and has been since refuted, evolution, finding out that global warming isn't a new-found hippy idea, and so on).

Rather than admit that scientists are actually their intellectual superiors and that they're only a guy that writes amusing books, they would rather rather say "fuck this, I'll do this my way and the ivory tower nerds can fuck themselves".

[McC]

Another aspect to hostility for science among some writers I think I have just realised: <snip>

...I'm pretty sure that if any published authors hold such a view, they are in the minority. I imagine most "writers" holding this view are probably forum warriors that have never actually been published, and with good reason.

[Junghalli]

A lot of times I'm not sure it's so much that writers really disdain writing hard science as that they tend to naturally take the path of least resistance, which means following in paths that are well-trodden by other writers that they've read, and in the case of sci fi this means following the lead of space opera authors who aren't particularly concerned with sticking to hard science, because that's what the majority of science fiction amounts to. Sci fi authors just starting out tend to gravitate toward space opera stories with FTL and galactic empires or WWII in space because that's what most other people are writing, and it's easiest just to follow what everybody else is doing. Hard SF requires you to consider issues most SF handwaves away, and so requires more originality, thus it's harder to write. Even in areas where it's not necessarily harder than soft SF it's harder in practice because you have to do more of your work from scratch. It's easier to write space battles by mixing together ideas from Wing Commander, Honor Harrington, and Star Trek than to do it by thinking about what real space battles might actually be like, which requires learning some basic science and extrapolating from that.

I know when I started writing I just sort of naturally gravitated toward writing basically a generic B5/Star Trek like universe because that was comfortable territory to me, it was only later that I started to work up the balls to break out of the conventions and do something more realistic and original. I imagine authors who didn't start out with a clear idea of wanting to write relatively hard SF like I did might often never feel sufficient incentive to take that step out of their comfort zone. As most writers feel inspired to go into a certain kind of SF by the works of other authors so when they become authors themselves they in turn help push others toward that same kind of SF, and the trend is self-reinforcing.

Of course, these same writers often tend to feel criticized and attacked when the hard SF people who do exist point out that their works are unrealistic, hence a lot of the defensive "the scientific hardness doesn't really matter" statements you'll see out there.

[Simon_Jester]

The industrial robots are designed to interface with human operators socially, in the sense that you can walk up to one and tell it what to do, not just through a programming language.

No, the robots were getting to the point where they were programmed to perform a specific task and nothing else- in Lenny Calvin notes that they were made to only mine one kind of mineral, not be general robots and she wants to make ones that can learn. So they would have social programming, but they wouldn't form friendships- that would be an application of learning.

Maybe I wasn't clear; I view the robots' strange social behavior as a sign that some underlying bug is causing them to issue scrambled social messages. So the robot runs around in eight-mile circles spouting Gilbert and Sullivan or founds a religion dedicated to the complicated piece of industrial machinery it maintains or whatever because part of its programming gives it social cues that can be turned into miscues by a bug.

Of course, later in the series we do see robots becoming increasingly single-purposed rather than general-purpose, though they still retain the ability to understand informal speech and such. This may be a reaction to the number of incidents where robots acted in 'social strange' ways in the past.
______

Though I don't think Asimov pulled that solution out of his hat; it makes at least some sense that a robot would start acting strangely and buggily if it was being overloaded by having to handle too many subordinate devices.

It makes no sense. If a robot is being overloaded by information the most likely responce is to reduce the amount of input, not increase it.

I think the problem was that the robot was not programmed to shut down one of its own control channels. It had to do something, and once it found itself with more questions to answer than it could address per unit time it flipped out. Things like this actually happened to some computers back in the early days; I seem to recall the computer on one of the lunar landers locking down because it was getting more input per second than it could process in a second. The robot design is stupid, yes, but that's why it got caught by the alpha/beta testers and not the end users.

That wasn't the point of The Naked Sun. The point was that having handicapped robots is a problem because other people will not have such inhibitions and will instead design different robots. You can accept Bailey's explanation or you can realize it makes alot more sense if the robots that the villian constructed were different. After all, handing your arm over to a women in a screaming rage is not something I can see justified by the second law.

OK, yes, you're right. I suspect that the Three Laws design was intended more as a way for US Robots, specifically, to reassure its customers than anything else. It does work for the robots you build; if it doesn't work for other robots designed by other people, that's a problem, but at least it's not your direct responsibility. I don't remember any laws requiring ALL robots to have the laws, after all.

They're somewhat more effective when the robots are restricted to dangerous environments and are not given access to resources they could use to take over civilization just by having lots of computing power- Asimov did not predict the Internet, for instance. Likewise for imposing hardware limitations (circuitry that can't support a supermind) and software limitations: once you've got an AI that is "smart enough" for the environment you actually want to put it in, you don't wire it for continued self-reprogramming.

The Evitable Conflict has the minds which are deliberately manipulating human society to make it optinum.

Yes. I'm not saying the Three Laws were a perfect safeguard. I'm saying they are a better-than-nothing safeguard, and I think Asimov did a very good job of showing how they can break down in several possible ways- and illustrating how computers tend to give you what you ask for and not what you want.

requires that robots stand down for reprogramming and maintenance even if that undermines their brilliant scheme for protecting all humans forever by sticking them in jars, and so on.

Robots are maintained by other robots. And in fact that is the situation on Solaria.

True, but at this point you're conflating multiple points in the 'Robot era'. I'm talking about the 'US Robots' period. After people overcame the fear of robots, they started trusting the robots more... until the circumstances that made the Three Laws work for roughly a hundred years broke down because robots were being used in roles US Robots hadn't been building them for.