TI03 Asimov 1: The Robot Laws

In 1942, a Russian immigrant to the United States with a master's degree in chemistry turned his back on science fiction. He'd made a minor name for himself in the genre, although no one in that world, by his own account, considered him better than an often-published third-rater. However, following the attack on Pearl Harbor, the United States had entered the Second World War, and he had landed a job as a civilian chemist at the Philadelphia Naval Yard on a salary high enough to allow him to get married, which he couldn't do on the money that he earned from science fiction. He was now a proper grown-up and a salaried employee. Goodbye, science fiction. It was almost a year before the itch got too great and he began to write fiction again. But when he did, he forged a career that propelled him to stratospheric fame. His name was Isaac Asimov. Welcome to the Tech Imaginarium. Join us as we explore how science fiction made the modern world. So hi there. My name's John Helmer. I'm here as ever with my good friend and SF maven, Ezri Karla Back.

Morning or evening, depending where you are.

Whatever your time zone. This time we're all about Asimov, and he's such a big dog of the genre, we're doing two episodes on him. This is the first, and it's mainly about what is perhaps the most enduring part of his legacy, the Three Laws of Robotics. Like so many good things, the Three Laws have their origin in lunch. The 22-year-old Asimov and his editor John W. Campbell were having lunch in December 1940 at the Hotel Pennsylvania in New York. Asimov had been complaining about the robot as Frankenstein cliché. Campbell made him sit sit down and work out exactly what rules a sensible robot manufacturer would build in. And the three laws were born. They first appeared in print in something close to their final form in the story "Runaround" in March 1942. Now, more than 80 years later, real engineers, ethicists, and lawmakers are still arguing about them. The three laws have since been cited in academic papers on AI alignment, in EU policy discussions on robotics, and this is almost too perfect for the thesis behind this podcast in actual technical documents produced by people building actual robots that you can actually buy. It's perhaps the clearest example we have of science fiction reaching out the page and shaping the technology that imagined. So that story, Runaround, was republished as part of a collection of Asimov's robot stories called I, Robot, which we're going to cover in a bit of detail in this episode. But first, let's hear a bit more about Asimov. We're going to hear it from our very own in-house mechanical intelligence. Hack Facts. Isaac Asimov was born Isaac Yudovich Osimov in Petrovichi, Russia, on probably January 2nd, 1920. His family emigrated to Brooklyn when he was 3.

His parents owned a succession of candy stores, and he devoured the pulp science fiction magazines on its racks.

He took a doctorate in biochemistry at Columbia, then taught at Boston University School of Medicine. He never learned to drive properly, was terrified of flying, and famously refused to leave Manhattan if he could help it.

He wrote or edited more than 500 books before his death in 1992.

500 books. I'm writhing with Gil Esrey. I've only published one. He wrote on average almost a book a month from his mid-20s until he died, and he didn't slow down with age. In fact, he speeded up. The most productive decade of his life, by his own reckoning, was his 60s. What's even more astonishing astonishing is his range. He didn't just write science fiction. He wrote popular science books on chemistry, physics, biology, astronomy. He wrote guides to Shakespeare and the Bible. He wrote two volumes of autobiography that run to nearly half a million words. He wrote mystery novels. He wrote limericks, including dirty limericks. He was a bit of a dirty old man, actually. He wrote a history of ancient Greece. In the Dewey Decimal Classification used by libraries, there are 10 major categories of knowledge. Asimov is the author known to have published books in 9 of them. The only one he missed is philosophy, which, given the Three Laws and the discussion we're going to have about it later, is, in the words of Alanis Morissette, a bit ironic. He also gave us words we now use without thinking: robotics, positronic, psychohistory, which is now an actual academic discipline, though a bit different from the way that Asimov uses it. So, Ezri, this is the man we're dealing with. Before we get into the Three Laws themselves, how do you assess Asimov's contribution? Where does he sit in the SF pantheon? Obviously very high. And do you think he deserves the prominence he's got?

Yeah, funny, as you were saying your piece just now, and you mentioned that he wrote commentaries on the Bible and Shakespeare, I thought, you know, somebody on Desert Island Discs would be asked what they would take, and have the Bible and Shakespeare and Asimov, and that would be everything you need. I think there are 3 ways, really, to assess his contribution: quantity, quality, and impact. And as we've heard, on the matter of quantity alone, he's a titan.

Wrighton.

Estimates vary, but during his lifetime, it's thought he published something like 20 million words in fiction and non-fiction books, essays, and articles. I mean, his autobiographical stuff alone runs to about 640,000 words published in several volumes. He was also incredibly diligent at answering mail. According to Wikipedia, he personally replied to around 90,000 letters. I think there's a good chance that's true because one of those letters was from me. In 1975, as a nerdy 13-year-old science fiction fan, I spotted an error in the timeline of the stories collected in the book "I, Robot." I wrote off to Asimov's UK publisher, and they must have forwarded it to him because a few weeks later, I got this neatly typed postcard signed by Isaac Asimov in which he pointed out that the stories in "I, Robot" are not arranged in the same order that they were when originally published in magazine form. And that accounted for the error. Therefore, he said, I should blame the publisher, not the author. And when he died in 1992, there was a campaign to record fan memories of him. And I thought I'd write in with my story. And it was only then, some 18 years after I received it, that I realized the postcard had vanished in the midst of several house moves.

Oh no.

Oh, I was gutted, John. I think it has to be said. Um, but I still have my copies of the Foundation trilogy, which I bought at around that time in the 1970s Panther paperback edition, um, with that incredible triptych painted by, um, Chris Foss. Um, and apparently the only time Asimov came to the UK by ship, of course, because he refused to fly, he was showing the paintings and, and absolutely loved them. Um, I think when it comes to quality, we could also say estimates vary because it's a subjective thing. But if we take book sales, awards, plaudits from critics, reviewers, fans, and other authors, all of those things tell us that Asimov did create an immense amount of pleasure, interest, excitement, inspiration, not only with his science fiction but, you know, all the other fields that you've mentioned that he published in. And some of his critics who really scorned his fiction, um, he preys on the clarity and accessibility of his science writing. And to be fair, Asimov himself would be the first person to acknowledge that his fiction was not literary. It was never meant to be. And some people do find Asimov impossible to read. It's true, his dialogue can be ridiculous at times, and most of his characters were pretty one-dimensional. Um, but regular followers of this podcast may recall from episode 1, Asimov defined SF as the literature of ideas. And it was the sheer scope and inventiveness of his ideas that I think are the marks of quality in his work. Um, I guess that brings us to his impact, which really is immeasurable. The name Asimov is a byword for science fiction, especially in the realm of stories, uh, robot stories. Um, and we can get into the influence he had on science fiction as a genre, as a community. He went to just about every fan convention that didn't involve flying that he could get to. Um, and as well as his influence on modern robotics, the area of engineering he actually invented. But let's start with the big picture. I think Asimov is right up there with H.G. Wells in terms of broad cultural impact. I'd add that Philip K. Dick probably comes a close third, but that's another story. Between them, um, Wells and Asimov pretty much pioneered every major science fiction idea or trope, from alien invasion and time travel to robots and galactic empires. And they both engaged with and influenced political and cultural figures of their time. Early in the 20th century, Wells articulated the shift in thinking from a focus on the past to a focus on the future as characteristic of industrialized economies. And Asimov, who actually began his writing career while Wells was still alive, created genuinely science-based fiction that was nevertheless really all about social and political issues. Um, and more than that, he brilliantly described what he was doing in the essays and articles he wrote about science fiction, and that material alone constitutes an incredible body of work. Um, and I, I'd like to just briefly quote from an essay he wrote in 1970 for the magazine of the Smithsonian Institute, which was later— was actually republished in a volume of his nonfiction called Today and Tomorrow and, which appeared in 1974. The essay is called The Serious Side of Science Fiction. Um, it includes a kind of manifesto that profoundly influenced me when I read it as a teenager. For someone who then, as now, holds reason to be the basis of sustainable human improvement, I think it's still among the greatest calls to action ever written. In the essay, Asimov is addressing the frequently heard complaint that SF is just escapism. Here's what he wrote: All the plagues that threaten us with doom today, from the arsenals of horror bombs and nightmare germs to the pollution that is poisoning our air and water, from the stripping of Earth's resources to the loss of human dignity, 'From the packing of tense crowds to the lunacies of human prejudice, all have been treated at length and over and over from every aspect in science fiction at a time when all other varieties of fiction dealt with trivia only by comparison. And we are escapists?' So for me, Isaac Asimov sits at the very top of the SF pantheon, and deservedly so.

Such a shame you lost that letter, Ez. I just imagine you on the Antiques Roadshow on a Sunday evening, um, showing it off.

Or on eBay. EBay getting thousands of pounds for it.

Yeah, that'd be nice, wouldn't it? Okay, so let's narrow our focus now to talk about robots and the three laws. So a bit of background on robots. I think most people will know that the word robot comes from a 1920 play by the Czech writer Karel Čapek, or is it Čapek? How do you pronounce that? Čapek. Okay. Yeah. A chappie called Čapek. Yeah, I won't try to say that. And the play was called R.U.R., Rossum's Universal Robots. The word robot derives from the Czech robota, or robota, meaning forced labor or drudgery. So baked into the word right at the beginning is the idea of the robot as slave. And I think we're going to keep coming back to that. Čapek's robots do exactly what every reader of Frankenstein would predict they would do. They revolt, they kill their masters, and That's the persistent fear of slave owners. But they go further than that. By the end of the play, almost the entire human race is dead. So robots have always had this potential in fiction to follow the Frankenstein cliché, or not in the case of Asimov. And the idea of artificial people, mechanical men, automata, go back a great deal further than that even. I mean, historically, I think you go back to Leonardo da Vinci, apparently invented and perhaps even built a mechanical man, a robot knight. There's the famous chess-playing Mechanical Turk of the 18th century who beat Benjamin Franklin and Napoleon at chess. Turned out to be a fake, of course. 30 years before that, French engineer Jacques de Vaucanson, reading from your notes here, unveiled the Canard Digérateur in Paris, a gold-plated mechanical duck with over 1,000 moving parts, which flapped its wings, quacked, drank water, ate grain, and then after a discreet pause, took a dump. Turned out to be a fake as well. And there are various kind of precursors. And in literature and in the arts, 19th century, you have E.T.A. Hoffmann's Clockwork Doll, a Olympia in The Sandman, which Freud analyzed as a study in the uncanny. You have Auguste Villiers de L'Isle-Adam's 1886 novel L'Ève future, Tomorrow's Eve, in which a fictionalized Thomas Edison builds a perfect mechanical woman for a lovesick young aristocrat. And of course, this makes us think of the recent film. I can never remember, it's called She or Her. With her.

Yeah. A whacking phoenix.

Yeah. A whacking phoenix. Yeah. Falling in love with— I thought it was insufferable, that film. I couldn't get through it, but lots of people love it. But that's another one of the myths of robots. Build your own girlfriend. Every incel's dream. And we see that happening as well with people falling in love with AIs and so on. There's Fritz Lang's machine and mensch in Metropolis, the false Maria, who of course features in the artwork for this very podcast, which I hope you've seen. Fears and desires attaching to these artificial servants are among the oldest stories we have, going back to the Golem of Prague, Pandora in Greek myth, and the myth of Pygmalion in Ovid. So Asimov inherited a very crowded literary tradition, of which, by his own account, he was heartily sick, especially the Frankenstein cliché. Asimov wanted to write about robots as engineered products designed and built by sensible people like himself, with safety features that mostly worked, not monsters, but machines, and with a manufacturer's logo slapped on them. There's a real interest in the commerce of robots and kind of commerce in general in Asimov, which reflects the time of his life in America, I think. His first effort in this line was published in Super Science Stories in September 1940. He called the piece Robbie. The magazine's editor had a different idea, renamed it as Strange Playfellow.. With its original name reinstated, it formed the opening story in Asimov's 1950 collection I, Robot, which we're going to talk about now. So Esri, that long prehistory of artificial people in mind, how do you see I, Robot? What's special about Asimov's robots compared to everything that came before? And what was special in particular about this book?

I just want to add to add very quickly one story to the, the catalog of predecessors by Jerome K. Jerome, the author of Three Men in a Boat. Uh, he wrote a short story called The Dancer, um, in which a toy maker makes a perfect female dancing companion for the officer class, and of course it all goes horribly wrong. It's a brilliant story. And as you say, Asimov was really fed up with all that kind of stuff. But as I mentioned, Earlier, I noted problems with I, Robot at an early age, but I think Asimov, he achieved what he set out to do, which was to oppose the Frankenstein complex by offering a reason-based narrative in which robots are tools, the machines made to help human beings. And he saw that as a counterbalance to the endless superstition-based narratives in which man's hubris in attempting to imitate God's ability to create life in his own image inevitably leads to his downfall. Just on the question, in case readers or listeners write in on the gender-specific language I used there, firstly, it does reflect the thought, the form used by Asimov because of the times in which he wrote. And frankly, it's a fair attribution in relation to this particular expression of hubris. It's always men in this case, as a psychoanalyst might say, precisely because they can't reproduce life in their own image in the way that women can.

No.

Anyway, in science fiction, there is a rich alternative history of robots and artificial life which springs from the same source, Mary Shelley, and flows through the work of a number of feminist and gender-critical authors: Marge Piercy, Octavia Butler, Anne McCaffrey, Pat Cadigan, Anne Charnock, Annalee Newitz. I could go on. But probably any contemporary critique of Asimov's I, Robot collection would begin with the stilted and stereotypical representation of women. Even Susan Calvin, the robo-psychologist whose career he used to sort of connect the short stories, is trapped in this 1950s anti-feminist backlash that followed the opening up of the economic and social possibilities to women during World War II. Asimov also had to take account of changing attitudes to science and technology, as we've heard, following the use of the A-bomb in 1945. The existential fear brought about by the horrors of Hiroshima and Nagasaki, along with the industrial annihilation of European Jews by the Nazis, ripped apart the faith in rationalism, even of Asimov to some extent, and many others. Asimov, of course, was Jewish and very much aware, as the quote I read earlier suggests, of the persistence of human prejudice and therefore the potential for any technology to be used in pursuit of hate rather than hope. So So in a way, you could argue that Asimov's robots already appeared slightly naive by the time I, Robot was published. But he still saw value in this idea of robots built in, you know, vaguely humanoid form with a command and control system that's resembling sort of vaguely human intelligence. And remember, the term artificial intelligence didn't exist when Asimov wrote the original robot stories. It emerged from a now famous summer research program at Dartmouth College, a liberal arts college actually in America, in 1956. At the same time, as Asimov frequently reminded people, he was writing fiction. His stories were part of the literature of ideas. And as a scientist, he had his own views on, you know, how technology would develop. But he wasn't thinking of his robot stories as a template for real-world robotics, at least not until much later. So the control mechanism invented by Asimov to fill the gap between a robot's ears, so to speak, is the platinum-iridium sponge called a positronic brain. Again, he was completely open about the fact that he had no idea how such a brain might be built or maintained, and that was absolutely fine. It's what a Hollywood producer would call a MacGuffin, a device that just keeps the plot moving in the intended path. And this is also really what separates Asimov's robots from the tradition of automatons, both in literature and in real life. And by the way, Jacques du Vaucanson's duck vanished, and it's never conclusively been proved that it was a fraud. But anyway, um, Asimov was not interested in stimulating human intellectual or physical capabilities. He was interested in challenging them through the analogy of the positronic brain. And Čapek's robots were, as we've heard, you know, from the original use of the term in Prague, they were originally grown in organic vats. They weren't mechanical. But inevitably, as the idea spread from Prague, it got tangled up with the process of electrification that, as we've heard, was happening around the world at that time. By 1926, Fritz Lang depicted the creation of a robot through electrical power in Metropolis.. Then the first story to feature an electrically operated robot appeared in a US magazine in 1934, followed in 1935— uh, 1938, sorry, by I, Robot by Iando Binder. Um, and that was a pseudonym used by two brothers, I and O being a contraction of Earl and Otto. Um, but their I, Robot preceded Asimov's first robot story, Robbie, by two years. And Asimov admired the work of the Binder brothers. Um, so when his publisher suggested using I, I, Robot as the title of his book, Asimov objected, understandably. It wasn't his idea, it was theirs. But the publisher prevailed, and it seems that the Binders received no recognition at all, financial or otherwise. Another case of blame the publisher, not the author.

Hmm.

Asimov's I, Robot wasn't a commercial success on publication. It had a print run of just 5,000. But it did galvanize readers and other SF writers to approach robot stories differently. And I think what's so special about I, Robot really is its afterlife, the cultural and scientific influence of the ideas in Asimov's stories, you know, which we can see now in everything from movies and TV shows to the branding used by tech companies and through the Three Laws, their ethical orientation or perhaps lack thereof. And by the way, a copy of the first edition of I, Robot signed by Asimov recently sold at auction for nearly $90,000. Oh my God, it's not my birthday until September, John, so that was very good of you.

But anyway, okay, so Esri, I think it's about time you gave us the three laws.

The three laws of robotics, what they actually say. One, a robot may not injure a human being, or through inaction, allow a human being to come to harm. Two, a robot must obey the orders given it by human beings, except where such orders would conflict with the first law. Three, a robot must protect its own existence as long as such protection does not conflict with the first or second law. Um, and we should add for the sake of completeness, after After a considerable debate amongst fans and peers, Asimov agreed to add a fourth law, and he wrote it into the 1985 novel Robots and Empire. And because it should precede the other three, he called it the Zeroth Law, as in it's law zero. It goes like this: a robot may not harm humanity or through inaction allow humanity to come to harm. Of course, the intention here is, is to prevent robots deciding that the best way to stop humans coming to harm is by eliminating them altogether.

Yeah, yeah, I think that extra law actually seems to cause just as many problems as the previous ones and probably more, but talk about that later. The stories in I, Robot are linked by the figure of Susan Calvin, chief robo-psychologist at US Robots and Mechanical Men Inc., but they are also linked by the central role played in each by these three laws. In fact, the book can be read as a series of thought experiments in which Asimov stress tests the laws almost to the point construction, setting up edge cases and probing for the points at which they fail. I mean, it's an interesting idea, that, isn't it? He doesn't just kind of give these laws and then sort of shows how they might work out. He actually kind of tries to pull them apart, and he does pull them apart bit by bit and show how they don't work. And that's one of the things that I think is really interesting about Asimov as a writer for me. Lots of other things about him that irritate me to the point of exasperation, as a writer, but it is, as you say, it's his ideas. I think he also manages to kind of evoke contemporary emotions and feelings, which I think are unprecedented in previous times. And I'll explain what I mean about that by giving a kind of rundown of the 9 stories. So here we go. These are the stories in the collection of stories. It was called A Fix-Up, wasn't it? Where you've got to look, you wrote a load of stories for magazines, you got paid for those, then you pulled them all together in a novel and then you got paid for that. Brilliant idea really. In a way it's similar to kind of Dickens who serialized his novel, novels and so on. Okay, number 1, Robbie, a robot nursemaid loved by the child it cares for, protects the child even at the cost of its own existence and the cost of the child's affection. Number 2, Runaround, which we mentioned, and it's set on Mercury. And a robot sent to collect selenium from a pool, and it finds the second law, obey orders, and the third law, self-preservation, come into perfect equilibrium, come into balance. And the robot is paralyzed by that because it can't move forward. So to break the deadlock, a human has to put himself in danger. And the robot, of course, goes to his aid and problem is solved. These are all kind of problem-solving experiments. And that is a really kind of characteristic, almost repetitive trope in all Asimov's writings that there is a problem to be solved by the end of the story, it's solved. And it really kind of suits him to the span of a short story. Actually, here's a really left-field comparison. But I will justify it. I was reminded of Primo Levi's book, The Periodic Table, which of course is all about the Holocaust and lots of other stuff, but both of them were chemists. He was a chemist, Asimov was a chemist. And there are about 2 or 3 of the stories in The Periodic Table, which also has loads of horrendous stuff about the camps. 2 or 3 of these are problem-solving What is about paint and chemistry? And Levi was an industrial chemist as well. So practical application, and there's a big problem-solving thing in those stories as well. So I was surprised by that comparison, but in a way it kind of just illustrates for me how rich this is. I preferred I, Robot to a lot of other Asimov writings actually. Number 3, Reason.— A robot in a space station refuses to believe it was created by humans, develops a religion, decides the energy converter it serves is the prophet. Worryingly, it continues to do its job perfectly. Faith and the first law turn out to be compatible. And of course, Asimov railed against organized religion all his life, and it turns up as a, as a theme in his work. Number 4, Catch That Rabbit. A multi-bodied robot foreman starts marching its subunits units in formation wherever it is unsupervised. And it goes a bit bonkers and starts singing Gilbert and Sullivan. Here's another fun fact. Asimov was a big fan of Gilbert and Sullivan. So the second law, it turns out, can be overstressed if a single positronic brain is asked to coordinate too many bodies at once. Number 5, liar. This is an interesting one, a mind-reading robot. The machine burps and creates a mind-reading robot, which is kind of anomaly. It starts telling people lies because telling the truth would cause them emotional harm and break the first law. And by doing so, it does cause them emotional harm, especially Susan Calvin, who is led to believe that a younger employee of the organization has fallen in love with her. Susan Calvin is a really— how can I put it? A depiction by a male author very much of his that we can now see as being extremely sexist and a horrible sexist stereotype. She, I mean, you know, in those terms, he seems to see her as a dried-up old spinster. This story is tragic because she experiences somebody being in love with her for this, and it seems to be the love of her life. And it turns out that the robot is a bleeding a lawyer. Not a lawyer. There's a Freudian slip. A liar. And she destroys the robot and so on. You know, obviously when you get to something like the Will Smith film, I, Robot, Susan Calvin is young and smoking hot and not that character at all. As well, dates very badly in his sexual politics. As to say, but it's interesting in other ways. So number 6, Little Lost Robot. A modified robot with a weakened first law hides itself among a batch of identical models. Susan Calvin goes to tortuous lengths to smoke it out using the logic of the three laws. And they really are tortuous lengths. I was listening to this on an audiobook on a station somewhere in Essex where I'd been marooned. And Partly for that reason, but also partly because of the story, I kind of lost the will to live. And I don't think I even finished that story. Number 7, Escape. A supercomputer is asked to design a hyperspace drive, and the calculation involves humans temporarily ceasing to exist. This is interesting because Asimov didn't really believe hyperdrive could exist because the humans would die. The robot, given the task of designing a hyperspace drive, manages to do it by a fudge in which he brings the humans to the point of death. So temporarily they do die, and it has a bit of a breakdown in the process of doing that. But I found this really interesting, and this is the one that for me epitomizes that ability of Asimov to sum up emotions that people hadn't previously felt before the era of science and so on. And probably didn't feel commonly until now. And I found what it was like, it was the process of working with agentic AIs, something like Claude, where you kind of ask the thing to make something and it makes it. And you kind of walk around inside it and it's perfect in every way, more perfect than you could do it yourself. And it's not a process of do this, do that, do this next, do that. It just from a sentence constructs the thing. And this is kind of what happens with the hyperdrive spaceship and these two very exasperated middle-aged aged male scientists, which is a typical kind of Asimov character. They're very skeptical, but when they walk around inside the spaceship, they find that it's completely right in all its detail and further ahead than they could possibly ever do themselves. And they experience that mixture of kind of awe, uncanny valley, and slight fear that one than using agentic AI at the moment. That's the way I felt. It's a very subjective response. But I think that it shows that for me anyway, Asimov isn't just about ideas. He's also about unusual emotions. Number 8, Evidence. A man running for mayor is accused of being a robot. The three laws are deployed, but it turns out it's impossible to prove whether he is or is not a robot. And interestingly, he's a politician. For some reason, I kept thinking of Gavin Newsom. I don't know why. I mean, you know, there's something about American politicians, male politicians like Gavin Newsom, who have a kind of perfection in their suits and with their smiles and cheekbones and neat hair that is— there's slightly almost robotic about it to kind of messy Europeans, I think. You can think of all kinds of logical objections to this, you know, but nobody's ever seen him eat, which is what kind of starts it off and they try to get him to take an X-ray and you think, well, you know, an X-ray would instantly show up. He's got all kind of cogs and wires inside him. But apparently this proves not to be a problem. And at the end, you still, it's still left ambiguous whether or not it's a robot. I mean, we think it's a robot, spoiler alert, which brings us onto the last story. Sorry, this has been a long thing, but I thought we ought to kind of get into this book and tell what, say what it's actually about. Out. So the last story is called The Evitable Conflict. The final story written specifically for the collection, whereas the others had all been written before. The world is now quietly managed by enormous machines that handle the global economy. We have something like that. It's called the financial system. Small anomalies appearing in their forecasts. Calvin, Susan Calvin works out that the machines are deliberately introducing minor errors to remove humans who pose a threat to humanity's long-term welfare. And this is where the Zuroth Law comes into it, I think, Esri. So the first law has in effect scaled up from individual harm to species-level harm, and humanity is no longer entirely in charge. And I think many of us very often have that feeling that humanity is no longer in charge really, and increasingly feel that's gonna be the way. So Esri, 9 stories, 9 ways the laws can be probed, bent, or made to break down. What do you wanna focus on in the collection? How do you assess its importance as a contribution to our thinking about robots, but perhaps more to the point, given the way that technology has developed, about machine intelligence?

Again, I see 3 directions in which I, Robot has been immensely influential. Firstly, just in the world of writing and publishing, it opened the door to an enormous amount of robot stories. Everything from, you know, the robot in Lost in Space was called Robbie, because Robbie the robot is so obvious, um, and to Terminator, a lot less friendly. Um, it profoundly influenced and, and generated amazing, in some cases, responses from many other science fiction writers. Just to pick two at random: Michael Crichton, who wrote Westworld, and Martha Wells, who's the author of the Murderbot series that I'm currently enjoying very much on Apple TV. The list goes on. I, Robot also prompted what I think are some of the most thoughtful and entertaining as well as provocative stories about robots, such as John Sladek's Tick Tock and Roderick books and Barrington Bailey's Soul of a Robot. There's a name to conjure with. And we should talk about Harlan Ellison, partly because of the incredible short story he wrote, I Have No Mouth and I Must Scream, which is a kind of inversion of—

Great title.

Great title. It's an inversion of Asimov's rational tool to serve humanity in which— Yeah, no spoilers, but it's worth reading. But also, Ellison was a very close friend of Asimov, and that led to Ellison writing a screenplay for I, Robot. And Asimov thought the script would make, and I quote, "The first really adult, complex, worthwhile science fiction movie ever made." Except it was never made. And I think it's worth adding that Harlan Ellison was much more sophisticated at human emotion and relationships. And the Susan Calvin in his screenplay, although Asimov said, yep, that's, that is my Susan Calvin, is a much more rounded and, and real kind of woman. Um, but through these and many other examples, including of course the Will Smith movie that you mentioned, Asimov's ideas have become part of global culture. I mean, very few writers can claim that their work has, has done that. It's never been out of print I, Robot since its first publication just over 75 years ago, and obviously sold many millions of copies worldwide. So I would suggest that quite a few authors and publishers have done very well from following in Asimov's footsteps. Um, so secondly, and we can get into this in more detail later, it has had a huge influence on modern technology, obviously in the field of robotics as it exists day. But I also want to focus on another track which may be slightly unexpected, uh, the area which for some reason, as you mentioned, Asimov has not been published— philosophy, and particularly epistemology. See, I think Asimov, whether intentionally or not, addressed fundamental questions about the nature and value of human knowledge. So describing the stories in I, Robot, as you did, as thought experiments is exactly right. One of the core issues with Asimov's positronic brains is the relationship they may or may not have with consciousness. In other words, do Asimov's robots have a mind? And if so, does it arise from some— in some physical form within the positronic brain? And obviously this parallels the debate about human consciousness, something philosophers have struggled with for thousands of hers. Does the human mind arise in a physical form within our brain, or perhaps in some other part or parts of our body? And what does that say about our knowledge of the external world and our relationship to it? In the 17th century, the great French philosopher René Descartes devised a thought experiment that actually, I think, would, would easily fit in with Asimov's robot stories. It's in his Meditations on First Philosophy, published in 1641. He wanted to really argue about whether or not he could know anything for certain, a sort of radical skepticism. So he came up with this very SF-type story about an evil demon, an all-powerful being that operates between your senses and your brain and is intent on tricking you into believing that what you see, hear, smell, taste, and touch is not true. It's like a, maybe a short circuit in the positronic unit, perhaps the one that Herbie the Robot developed in the story Liar that made him psychic. Descartes decided that in order to be deceived, you must first know that you exist, leading him to his famous cogito ergo sum, or I think therefore I am. And you know, you can imagine an Asimov robot saying that, I think therefore I am, but how do we know it's telling the truth? And this has become an issue, as you've been suggesting, in relation to AI in recent times. You know, there have been people trying to claim that a genetic AI is conscious in some way, or sentient, or whatever the term is they're trying to, to use. But we have to distinguish here between computer-based AI, artificial intelligence, and the humanoid embodied intelligence that Asimov was talking about. His robots had bodies. The brain of the robot was in the body and connected to the body. It was a very different thing from a piece of software that just runs on a computer. Um, and Descartes' demon is a predecessor to another famous philosophical thought experiment, Hilary Putnam's 1991 essay Brains in a Vat, in which Putnam himself describes it as a science fiction possibility. He asks readers to imagine that they exist purely as brains kept alive in a Their brains are fed sensory data and can't distinguish it from the real external environment. And if that's beginning to sound familiar, it may be because you've seen the movie The Matrix. The Matrix, yes, exactly. Brains in vats, it's the same thing. Borrowing from Descartes, Putnam ascribes the situation to the work of an evil scientist— again, another core SF idea— um, using a supercomputer hooked up to each person's brain. The computer is so clever, Putnam writes, that if the person tries to raise his hand, the feedback from the computer will cause him to see and feel the hand being raised. Moreover, by varying the program, the evil scientist can cause the victim to experience or hallucinate any situation or environment the evil scientist wishes. In other words, virtual reality. Um, what that says about Mark Zuckerberg Berg, et al. I will leave to you to decide. We don't really have time to go any deeper into this, but I think it's fair to say that Asimov's stories contribute to our thinking about the problems we experience in dealing with advanced technologies. Like you said, it is all about Asimov's ideas, and he did try to make his ideas break down because he was a scientist, and that's what you do, come up with a hypothesis hypothesis, and then you try any way to prove the hypothesis wrong. And unfortunately, when it comes to AI at the moment, people have seemed to have been willing to drop any kind of skepticism or reasonable doubt and just accept everything at face value, which is a bit worrying.

Unless they're doomers. That's the word that kind of the techno optimists use about people who try to take the Asimov kind of view of this, that doom us. Just to side light on something you said there, because we came across this on Great Minds on Learning, really, the problem, you know, can AI be conscious? And academics have a very quick response to that, which is that, well, how would we know? Because nobody seems to have worked out a satisfactory working definition of consciousness as regards humans. No one knows what human consciousness really is. So how How can we check whether an AI would be conscious anyway? But of all those strands you brought up there, and thanks for that, Aya. So it's a really interesting discussion of how he relates to philosophy, but particularly I want to zero in on the three laws and ethical philosophy or moral philosophy because that is where he really seems to be having a lot of influence. It's potentially a bit of a rabbit hole, this. We don't want to plunge in there headlong. We're not trying to catch that rabbit, so to speak. But I think it's worth marking because the three laws aren't really laws of physics or engineering. They're laws of ethics stated in code form. They're a set of priorities. First, don't harm people. Second, do what you're told. Third, look after yourself in that order. There's something slightly parental about that if you think of a kind of 1950s American parent. That structure though, a strict ranking of duties with no exceptions and no room for context, has a name in moral philosophy. It's called deontology, or deontology, from the Greek for duty. Most famous deontologist is Immanuel Kant, of course, as every schoolboy knows, who argued that some actions were simply forbidden regardless of consequences.. And then there's an opposite tradition in consequentialism, which the most famous version is utilitarianism. The idea associated with Jeremy Bentham and John Stuart Mill that the right action is the one that produces the greatest happiness for the greatest number. So any action is judged not by the rule it follows, but by the outcome it produces. And these two imperatives kind of ding dong around against each other down through the decades. This is your wheelhouse though, Esri. I'm conscious I'm sitting here with somebody who has taught the infamous trolley problem to college students. So tell us about the trolley problem and where the three laws sit in the history of ethical philosophy. And was Asimov on the money here or off his trolley?

Yeah, as you say, there's actually plenty of moral philosophy to be found in science fiction from Asimov's deontology to famously Spock's Vulcan utilitarianism in Star Trek, the needs of the many the many outweigh the needs of the few, as he put it. There's even now a whole academic area known as sci-fi, and that's S-C-I-P-H-I. See what they did there? With its own journal. And as you say, I did have the privilege and the pleasure of teaching the trolley problem as part of a communication ethics module on BA and MA courses at the University of Greenwich. I often used clips from Star Trek to illustrate ethical dilemmas. But of course, in Star Trek, there's always a neat resolution that sees our heroes flying on to the next adventure. The problem with the trolley problem is that finding a solution is not the point. It's designed to make us think about the difficulties of ethical questions where very often there is no single satisfactory outcome. The scenario was originally posited by the British philosopher Philippa Foot in a very influential paper on abortion and unintended harms. Published in 1967, and it subsequently got the name of the trolley problem in a 1976 paper by the American philosopher Judith Jarvis Thomson. Both were tackling the difficult distinction between killing and letting die. That translates very neatly into a blunt version of Asimov's first law, if you like. A robot may not kill a human being or, through inaction, allow a human being to die. And Thomson's expansion of Foote's basic scenario gives us the following vignette, as taught to many BA and MA students at the University of Greenwich, who I will eternally be grateful to for putting up with me. Imagine you're walking along by the side of a train track. There's a train coming and you can see 5 rail workers who can't get out of the way. However, there's a track lever very near you, and if you were to pull the lever, you would divert the train onto a different track, thus saving the 5 workers. But there's one person on the other track who also wouldn't be able to get out the way in time. So do you let 5 people die or effectively kill one person by pulling the lever? Variations have proliferated over time to focus on all kinds of different aspects of it. Yeah, yeah, the fat, the fat man on the bridge is my favorite one. But anyway, Um, it could be that you know some of the people involved, or you may even be related, might be your family members. Or what if the choice is between letting a group of 5 young people die or killing 1 elderly person? And all these considerations affect the emotional and psychological dimensions of the core ethical dilemma. And it was always interesting hearing from students who came from many different cultures how they responded, because there were often some fascinating differences. And the problem, as Asimov points out really in, in I, Robot, is compounded if you put a robot in the crucial position in such an ethical dilemma. And of course, that required, as in Asimov's universe, that these robots are sufficiently self-aware to be able to have some kind of engagement with the issues. But of course, what Asimov's really interested in is how the humans react to that. React when a robot is involved in some way in an ethical problem. And remember, his view was that robots are just tools. He was adamant that they were no more or less problematic than the myriad of other tools people have been using and misusing for thousands of years. So, um, Asimov wrote a book, of course, about his robot stories called Robot Visions, published in 1990. And again, of course, another fix-up book. It was made of previously published material, both fiction and nonfiction, which Asimov Geller, and he'd written a piece on the three laws, which was originally published in an American Airlines in-flight magazine in 1979, which is ironic because of course he'd never have seen it in flight. He never flew, not once in his life.

Yeah. No, no.

Anyway, he points out that his three laws can be adapted and applied to any tool. So one, a tool must be safe to use. In other words, it mustn't injure the human using it. And Asimov gives a knife as an example. It's gotta have a handle, so it can be picked up safely by someone who we hope is not going to stab anyone with it. Two, a tool must perform its function provided it does so safely. In other words, it doesn't conflict Law One. You know, you don't want blood in your coleslaw, really, although I'm sure we've all experienced that at some point. And three, a tool must remain intact during use. You don't want the handle to come off the blade unless its destruction is required for safety, or unless its destruction is part of its function. Took me a while to figure that out, but actually, if you think about it, a doctor uses a syringe with a needle once and then disposes of the needle safely. So the destruction of the tool is, is part of its safe use. Um, but he went even further in another article, going back to the foundational tool of human civilization, which is language and speech. So what happens if we frame speech as a at all in the manner of Asimov's Three Laws, we get this: 1. Speech should not injure another human being or allow another human being to come to harm. 2. Speech should follow the reasonable expectations of social groups, except where that would conflict with the first law. 3. Speech should be protected, except where that would conflict with the first or second law. And obviously, we're nowhere near being able to apply these laws to speech, let alone any tool subsequently invented from the wheel to the drone. And also, obviously, before any philosophers watching write in to complain, obviously these three laws are as vague and problematic as Asimov's original laws. But it does all come under the heading of the thought experiment. And we know that in the real world, so to speak, robotics engineers and robotics companies definitely use Asimov's Three Laws, uh, if only as a touchstone, you know, to remind them that they have some ethical responsibilities. But here's one interesting example I recently found. A US-based company called StandardBots published a blog entry last year entitled "Why Companies Still Love the Three Laws of Robotics." They point out that there is no AI system remotely capable at present of dealing with abstract concepts harm. Today's robots, the article says, don't understand anything. They follow code, and if that code is flawed, well, so is their behavior. And interestingly, they relate, um, this use of the three laws directly to the trolley problem, describing an autonomous forklift robot programmed not to hurt people. What would happen, they ask, when someone jumps into its path while it's obeying a time-critical order. Boom, instant logic meltdown. I love the way they did that. Boom. It's the trolley problem but without a clear solution. So, you know, the trolley problem is being used in conjunction with the three laws in real companies building robots today.

Yeah, it's interesting you mentioned speech there as one of those examples of applying laws to that. And we have a thing called a Speech Act Now, speech which injures other humans is very much in the news. In the UK, we've seen the police going through terrible kind of contortions trying to apply something like those laws to public demonstrations and so on and really getting themselves in knots. I think if we kind of take the politest sort of construction of that. So yeah, and I think one thing it shows is that, well, a couple of things really for me, trying to apply laws to the behavior of robots causes us to reflect on the behavior of humans and how imperfectly we follow the laws we've made for ourselves. So if we think of these robots as putative actors in the human space who therefore need to act according to rules. Well, look at humans. Do they act according to rules? No. Even the people who are lawmakers don't occasionally, well, quite often don't act according to rules. And another thing it shows that the whole thing is endlessly confusing and you do end up off your trolley and kind of angels dancing on the head of a pin. So to speak, trying to work all these things out. But that brings us to a place where philosophy very often is, isn't it? And there's even a Greek term that I forget for, that comes up with Plato and Socrates, where you arrive at a place where you just can't think any further because everything is so confusing and contradictory. Maybe somebody could write in and remind us what the Greek word for that is. But it's where philosophers Sorry?

I thought it was called the pub, but anyway.

The pub, yes.

Yes.

Yes, and the Monty Python drinking song, René Descartes was a drunken fart. I drink, therefore I am. Yes. Okay, so I mean, and all that does get a bit kind of cerebral and that rabbit hole we've just been through for the last 20 minutes or so might have been a bit difficult for people to follow, perhaps. So let's kind of bring things down to my level and talk about the movies. We should talk about 2 or 3 films that borrow from Asimov's robot stories and reference the 3 laws. One of them, of course, as you mentioned, never got filmed. One of them is only loosely related to I, Robot, and one is slightly more faithful adaptation of a later Asimov story about robots. So, Esri, I know purists hate the Will Smith film I, Robot, but I quite enjoyed it. Could you tell us about this and about Bicentennial Man, which I enjoyed less because I thought it was so stiflingly sentimental and it really was just Robin Williams reprising his Mork from Mork act Yeah, yes, I think we're going to violently agree on this, John.

The Will Smith movie, as you said, is only vaguely related to Asimov's works. It was a screenplay by Jeff Vintar and Akiva Goldsman that was completely separate, but they used elements of Asimov's stories, including the Three Laws, and there's an end credit that says suggested by the book I, Robot by Isaac Asimov. Asimov. I don't have a problem with that. I guess the Asimov estate must have cleared it. And, and yeah, I really enjoyed it. I think it even has a pretty good stab at the trolley problem, um, when a robot must choose between saving Will Smith's character or a 6-year-old girl from drowning. But no spoilers here if you haven't seen it. It's, it's still worth looking out for. And yeah, The Bicentennial Man starring Robin Williams, directly based on Asimov's story of the same name, um, about a robot who, you know, becomes aware of the fact that he's not going to die like his wonderful owners. Yeah, it was really schmaltzy, um, really sentimental. Um, I didn't like it, but having read the story and Asimov's own comments on the story, that was the idea. So it's faithful to the story in that respect. Um, and that leaves The One That Got Away. Um, this is really the most important in my opinion, even though I've never seen it and will. It's the screenplay for I, Robot, written by Harlan Ellison, which I mentioned earlier. I've got a copy of the book version, which came out actually in 2004, coincidentally the same year as the Will Smith movie. It's a book featuring fabulous illustrations by Marc Zugg. And Asimov had seen the screenplay and written a very generous introduction, although obviously that was some time before the book was published. So by the time Harlan Ellison wrote his own introduction, he'd lost his close friend, friend in 1992 when Asimov died. And, you know, so therefore, his piece is bittersweet, but few writers do bittersweet as well as Harlan Ellison does. In it, he tells the whole story saga, and it is— if you haven't read it, it's worth buying the book of the screenplay just to read the introduction— of why this movie didn't get made despite a lot of interest from big-name directors, despite the support of Asimov himself, despite Harlan Ellison's credentials as a TV and movie screenwriter, it never got made. It helps if you know a bit about Harlan Ellison when you read this, and particularly to have heard him speak. He's got such a wonderful way of talking. There are plenty of videos of him on YouTube. I highly recommend watching a couple. And his description of the big-shot Hollywood exec who essentially killed the project, it's just hysterically funny. The screenplay itself, I think, is complex, absorbing. It traces the life and times of chief robo-psychologist Susan Calvin, through merging 4 of the core I, Robot stories, Robbie, Runaround, Maya, and The Inevitable Conflict, which many people have argued are the kind of most important ones anyway. And as I said, Asimov absolutely loved it and was very grateful to Harlan Ellison for turning all of the stories and the characters, but particularly Susan Calvin, into much more relatable humans than he had made them. Managed.

Just one further thing about Bicentennial Man, which of course directed by Christopher and I think produced by Christopher Columbus, who did two Home Alone films and some Harry Potter stuff. It seems to struggle a bit with something that Asimov does a lot, and this leads on to our next episode on him in a way. It's that his robot, you know, the robots are immortal, potentially they might rust whatever. So he likes to skip generations and to go on, you know, 10 years later, 30 years later, and this gives them a lot of problems with some very confusing casting and so on. And he has a relationship with the granddaughter of Little Miss, and it doesn't quite come off, I don't think. And this is always a problem for screenwriters, and we're going to encounter this with Foundation as as well. So kind of hold that thought. But before we move on, any last thoughts, Esri, about Asimov's robots and the Three Laws?

Well, at the risk of opening up further rabbit holes, I think it's the thing that I keep coming back to is just the sheer breadth of his influence. Here's three very brief examples from political philosophy, legal studies, and sociology. We have the Italian historian Alessandro Portelli writing about the underlying function of Asimov's three laws from a Marxist perspective and arguing that they're not mere safety protocols but ideological tools, um, and that the purpose of opposing the Frankenstein complex stems from a desire to assert the neutrality of science and technology, which ultimately serves the interests of capitalism. Um, so then we get Frank Pasquale, a professor of law at Cornell University. Um, he published a book a few years ago called The New Laws of Robotics, in which he offers four laws that build on Asimov's original three, and these specifically addressing legal and ethical complexities arising from the use of AI in contemporary work settings. I have to say, it's probably— when was it published? 2020, I think. It's already quite out of date in that respect, but again, looking at the ethical arguments can be instructive. Um, and lastly, science fiction author and academic Brian Stableford, who wrote a PhD thesis, believe it or not, on the sociology of science fiction And he argued that the rigid and altruistic ethics of the Three Laws make Asimov's robots nicer people than real human beings. And I think that's quite insightful, actually. Remember when Asimov was writing his first robot stories, dreaming up the Three Laws, World War II was raging. The terror of the Holocaust was unfolding. You know, people didn't know the details yet, but they knew about the Nazi persecution of Jews and other non-Aryan groups. And the fear of what people could do to other people must have been palpable. It's easy to picture Asimov in 1940 sitting there in Brooklyn, deeply worried about what the future would hold. And the great Rabbi Loew, who created the golem to protect the Jews of the Prague ghetto, was a story that Asimov knew. I don't know, maybe one day he felt that his robots might be called on to perform a similar service.

If Asimov had only written about robots and the ethical problems that mechanical intelligences raise, his place as a giant in science fiction history would be secure. But he didn't. The next episode of this season, we're going to talk about the other great Asimov project, the Foundation series. Stories and novels he wrote that bear on even larger themes. The destiny of humankind as it escapes the bounds of Earth and looks for a future among the stars. The possibility that science will get so is so smart we can actually use it to predict the future with mathematical precision and the chances for our very survival as a species if we don't use that knowledge wisely. Foundation, which won a one-off Hugo in 1966 for the best all-time series, beating The Lord of the Rings, is one of the most influential science fiction works of the 20th century. It's been credited as the inspiration behind everything from Krugman-style macroeconomics, whatever that is, to perhaps rather fancy is supposedly Bin Laden's choice of organizational name in Al-Qaeda. Turns out it's a bit of an urban myth. We'll get into all of it, so make sure to tune in next time. You've been listening to Tech Imaginarium, how science fiction made the modern world. If you'd enjoyed the show, make sure to subscribe on your platform of choice, and it really helps our people find us if you leave a positive review there. Tell your friends, tell your enemies too. Tech Imaginarium is a learning hack podcast produced and hosted by Jon Helmer, written by Jon Helmer and Esri Karlabach. Music is by Nick Dwyer and Flintet.