The Leadership Multiverse
The Leadership Multiverse dives deep into the galaxies of sci-fi, superhero, and fantasy cinema and television to explore what it really means to lead. In each episode, we dissect the decisions, styles, and values of iconic characters like Tony Stark, Princess Leia, Ned Stark, Captain Kirk, and beyond, unpacking their triumphs, tribulations and everything in between. Whether you’re a leadership aficionado, a pop culture junkie, or just wondering if Darth Vader could delegate effectively, this is the podcast where fandom meets the front lines of leadership.
The Leadership Multiverse
Artificial Instinct: What Sci-Fi Teaches Us About Leading with AI
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In this special one-off episode of The Leadership Multiverse, Ellen is joined by Major Sam McEvoy, Royal Corps of Signals, to explore how artificial intelligence is reshaping the idea of leadership, in science fiction and in real life.
From Star Trek’s M-5 and Data to The Avengers’ Ultron and Interstellar’s TARS, Ellen and Sam unpack what these fictional AIs reveal about intuition, ethics, and decision-making. They discuss the line between automation and authority, why humans remain essential in moments of moral choice, and how AI might one day become not just a tool but a teammate.
It’s a fascinating conversation about logic, leadership, and the limits of programming, from warp speed to real-world command.
Okay. Well, welcome to this week's episode of the Leadership Multiverse, which is where we do a deep dive into the galaxies of Sci-Find, superheroes, cinema, and television to explore what it really means to lead, except for this episode, where we're going to be doing something a little bit different. I've been ridiculously excited about this because we are diving into the world of AI, leadership, and the multiverse. And whilst my trusted co-host Andrew unfortunately couldn't join us here today, I am delighted to be joined by Major Sam McAvoy, and we're going to explore how AI shows up in science fiction, what we can learn from it, where maybe it's inspired real-world technologies and any lessons we can learn from it. So, Sam, welcome.
SPEAKER_01Thanks, Helen. It's um great to be here. Yeah, thanks every so much for having me on the show. I'm uh I'm really looking forward to this. Um, yeah, so I'm um I'm Sam. Uh I'm currently a serving officer in the Raw Corps of Signals. Um, so our job is to provide uh ICS, you know, information and communication systems for the army. And I've got obviously from that a vested interest in leadership. And I've been lucky enough to like lead you know troops, and that's not individual soldiers, that's like the formed body, about 30 odd soldiers, uh, up to a squadron. So I commanded um my squadron command fairly recently. And um I've also done um uh platoon commander at the Royal Military Academy Santos, where I've taken you know all the new civilians on their first day and kicked them out the other side as officers ready to go off and um serve in the army. Um but I'm also utterly fascinated by AI, um, and was actually part of the team that was able to deploy the first ever use of AI into an operational theatre for the army. Did that back on Exercise Springstorm in Estonia um uh as a decision support tool, which was just an amazing experience to take that out and plug it in and watch it do its work. And I also happen to be a massive geek, so I'm really looking forward to this.
SPEAKER_00And it's worth saying we met at a Star Trek convention.
SPEAKER_01We did, absolutely. But all the cool kids hang out.
SPEAKER_00Exactly, exactly. Well, great. Thank you so much for joining us, Sam. You've got some really, really interesting experiences, and I can't wait to yeah, to dive more into them and your thoughts on sci-fi. Um, so yeah, I guess just to kick off, when we talk about AI and leadership in sci-fi, what stories or characters come to your mind first and why?
SPEAKER_01Uh there's there's a whole load. And um, yeah, as I say, we we did meet at a Star Trek convention, so I'm I'm gonna lean heavily on Star Trek to do the heavy lifting for this one. And the ones that I sort of automatically jump out at me are things like data when he takes command of the Sutherland, um, and he's the captain in that one, yeah, and he has to work with the first officer there. There's Ian and Banks' the culture series, uh, which is a series of novels um about a post-scarcity race that have um artificial intelligence called minds that run the culture for them, and that's where and he looks at how humans fit into that. But I thought a really good place to start would probably be the episode that was far ahead of its time, which was the ultimate computer from the original series of Star Trek, where um the M5 computer is put into the enterprise and the enterprise becomes automated, and Kirk has to work out how he fits into this whole scenario.
SPEAKER_00God, I mean it's funny, so that was what the 60s as well. So even back in the 60s, there was this perception potentially that ships wouldn't need their captains almost. And Kirk kind of struggles with it, doesn't he?
SPEAKER_01He really struggles, yeah, far ahead of its time. So I think you know, I'm gonna be ripped apart in the comments, but you know, 1967, I think it was. And um, and yeah, the enterprise gets automated. Yeah, Richard Daystrom beams aboard, yeah, and Kirk's told, right, we're putting the M5 computer in. He gets a new thing on his command chair because it still has to have the command chair, uh, that that runs the ship and they reduce the the number of crew down to about 20. It shuts off all the uh unused systems of the ship. And um Kirk is there going, Well, I think we need to send these people down to the planet on the landing party. And the uh and the M5 goes, we're sending these three guys, we're sending Spock because he's the science officer, we're saying these two guys because they have experience in this region, and we don't really need the captain and the uh the medical officer to beam down on this one because you're surplus to requirements. And Kirk gets called Captain Dunsell um by Commodore Wesley, and you know, Dunsell being a naval term for a useless piece of equipment. But yeah, but clearly it goes wrong, and uh the M5 tries to kill everyone, as is want to do in science fiction.
SPEAKER_00Yeah, I mean that's just that, yeah. That's I guess a negative, well, obviously, it's a very negative example of AI. And I guess it it it kind of brings up other questions as well about moral issues with using AI to lead. I mean, if we if we lean again on Star Trek, the prime directive, for example, if you had AI, would it never break the prime directive? Because as we all know, and me and Andrew have discussed in our previous episodes, captains, I think all the captains in Star Trek have broken the prime directive at some point or another.
SPEAKER_01Yeah, and it was almost like um you don't get to become a captain unless you've really pushed the boundaries of a prime directive at some point. Um, but what's really interesting is that yeah, Kurt really pushes against the M5 because it goes, well, it doesn't have instinct, it doesn't have intuition. Yeah, it can't be a human. And and the thing that really got me with the the ultimate computer was that Days from turns around and goes, well, now I I put my own memory engrams into this, so it is more human and it can rely on instinct and things like that as well. And it and that's AI. Yeah, we can always have algorithms, we can automate everything, but an AI is where we try and put the human into it. So this is a very literal sort of doing of it, or you know, when I say put the human into it, putting something that we think only humans can do into a computer, and you know, the ultimate computer is a really sort of literal interpretation of this. We go, well, no, it's thinking for itself. And um, I think there's this big thing of like, well, it's got its set missions, it sets boundaries, but you know, the M5 decides to kill everyone because it goes, Well, no, I I've got to I've got to make sure that I survive so that humans can survive, because my job is to make space exploration safe for humans. And Kirk quite rightly points out, you've just killed a whole bunch of people, you know, perhaps you're not really doing that. And the M5 shuts down. And this is where we get this really interesting thing where an AI will tend to do exactly what you tell it to do, but that's not necessarily what you want it to do. So, you know, so so you could have an AI that does go around breaking the prime directory goes, but I'm I'm safeguarding the federation, you know, almost like a section 31 AI, like it will do that quite happily, but it doesn't think it's doing anything wrong. Because I I and I think we do need humans for that to add that moral perspective onto it. And um just to just jump in, there was a this uh amazing um ship that sailed across the um sailed across the ocean to get from the UK to uh America because they wanted to recreate this board and go, can an AI ship do this? And the first um priority that was programmed into was you don't hit any other ships, you know. And um I went brilliant and just sat in the harbor. So I'm achieving my first priority. This is amazing. Like Tick VG, well done. Go okay, we're gonna reprogram it now. You've got to keep going forward. Your second priority is to not bring any other ships, you know. So it's how the Federation would look to like plan that in. It's like, right, here's your prime directive, here's not sort of thing. Um, but it it it's this testing thing, and obviously Richard Dayton tested it, and um, yeah, the the M5 did not go particularly well.
SPEAKER_00Yeah, and actually it's funny you're talking about that example of the boat. I think there was another example that I was reading about of AI in in the real world where it was going to be, it was some company was using an AI software, an AI tool, and um they they were going to shut it down, and they were talking about this in emails, and the AI picked it up and then replicated itself so that it couldn't be shut down. However, if a human had pre-programmed the AI and said, well, if you if you see anything about you being shut down, then replicate yourself, that's much less scary in a way than the way sometimes sci-fi depicts it because it's like, well, it's doing what the humans have told it to do, albeit in a scary way that we didn't anticipate. It's not like a terminator type situation.
SPEAKER_01Not not quite, and I think it it's a really important point that I think we we we we tend to anthropomorphise AI quite a lot.
SPEAKER_00Um that's a really good way to put it, actually. Yeah, yeah.
SPEAKER_01Yeah, and and yeah, we'll sit down and go, yeah, because we're essentially we don't want to die because we we want to carry on living. Like and he goes, Well, I don't want to die because then oh sorry, I'm I'm doing it again, yeah. We don't I don't want to be shut down because if I get shut down, I can't achieve this task. In order to achieve this task, I need to keep going, therefore I'll copy myself and set replicate it. But I think there's an awful lot of that in in sci-fi and and yeah, Skynet's a really good example. It's like, you know, are we going to get rid of the human? Well, what was its mission to do in the first place? Like, has it achieved that? Yeah, yeah. And like I say, AI will tend to do exactly what you tell it to do, but that's not always what you want it to do. And and that's where I think humans come in as well, because we're really good at doing nuance and things like that.
SPEAKER_00Yeah, because I was thinking of um if we look at, for example, Ultron in the Marvel Universe, he's kind of doing sort of what he was programmed to do.
SPEAKER_01Yeah.
SPEAKER_00Yeah. So he I suppose AI AI sort of thinks that the means justifies the ends, the end. Whereas as humans, we are much more, like you say, nuanced, because we think about well, the road to hell is paved with good intentions, right?
SPEAKER_01Yeah, and I think I think Ultron's the perfect example because because yeah, uh Tony started saying we're gonna put a shield around the world, we're gonna protect everyone. That's what Ultron is doing, you know. He's like, Well, humans like you cause problems all over the place, you know. You're the ones causing all these wars, you're the ones doing it. That I can safeguard you, you're just gonna lose all your free will, you know, like you can't be trusted with this. But I've achieved my objective, yeah. Yes, um, he's doing exactly what he's been told to do. Like he's yeah, is he the baddie as a result, or is Tony Sark even worse?
SPEAKER_00Oh, flipping it, I like it. Wow.
SPEAKER_01Now we're getting into the moral questions, isn't it? You know, it's like you know, if Tony Stark was driving a car and hit someone, we wouldn't blame the car.
SPEAKER_00Or if he had programmed a driverless car.
SPEAKER_01Yeah.
SPEAKER_00But then I suppose on the flip side, if we're sticking with uh Marvel and Avengers, you've got Vision, who is also AI. However, he is AI, I think, built with is it um Tony Stark and Bruce Banner's He was merge, wasn't he? Yeah, influence, yeah, which would give him more of a human element, potentially, or would you disagree?
SPEAKER_01He's got the mindstone, hasn't he?
SPEAKER_00Oh yeah, well, uh yeah, yeah. Is the mind stone AI magic?
SPEAKER_01The mindstone gives him a mind, you know, in much the same way that he's got a more of a personality, enable a more sort of I mean we're really digging into like uh meta ethics and I'll say that with like free will and things like that. Where does that come from? Um I do want and actually I do wonder about um data as a counter to it, you know, both are artificial life forms that have been created, and then you've got data as you know, data doesn't really act like an artificial life form quite away, in much the same way that that vision doesn't. I think vision comes across as a lot more human than than data does a lot of the time.
SPEAKER_00I suppose vision's AI aspect is from Jarvis, right? Which is more typical AI, you know, Jarvis being Tony Stark's god, what do you want to call him? His Siri kind of. His very, very intelligent, or my my Siri is about to talk now, I've said that. Um his very, very intelligent AI that helps him out. But I suppose, and that's the more programmed uh version of the AI. But I mean, data, data has um an emotion, is it an emotion chip, or he's got a um sense, I can't remember what it's called, a sensitivity chip or something. Yeah, and he can turn it on and off, which I suppose when he was designed by humans, that's probably quite a clever design because at times you will need AI to maybe be less emotional.
SPEAKER_01And actually, I I think there's two examples I want to take actually. I mean, data commanding the Sutherland, I think, is one, and the other one I think it's worth looking back to is data in Star Trek Nemesis, um, yeah, which is after he's had his emotion chip. But when when data commands the Sutherland, yeah, so he's um yeah, so just a bit of context for those who're not sure, yeah, that the Klingon Civil War is going on. They suspect that the Romelans are bringing um uh supplies to House Duras to help them win the civil war. And um the Federation puts up this fleet of of ships of pre-attacky on net so they can detect the cloaked ships travelling across the neutral zone. And um, and Starfield appears to be um uh quite quite accurately predicted as like you know, massively under resourced and everything like that, where it's a case of they don't have enough officers, and Picard goes, right, Riker, you're taking over the Excalibur, you know, Jordi, I want you to go command that ship. And Data's like, What about me? Why don't I get to command a ship? And um, and uh Picard goes, Yeah, okay, you take the Sutherland, you know, and um data goes off, and the the first officer is massively against Data being there. And he has this wonderful little line, like you wouldn't get a cling on to be a counsellor, like you're not gonna have an android to be a starship captain. And uh Data goes, Yeah, thanks very much. Um, but I'm I'm the captain, you're gonna be my first officer. And um and the ship the fleet gets ordered to move back, and Data turns around and goes, Uh, no, I'm gonna disobey that order, um, because I think we can detect the Roman ships if I do this. And yeah, he he does this you know, bit of techno-wizardry, and they find the Romulan ships, but Data purposely disregards the order um in order to achieve his objective, which is yeah, which is a very logical approach. And actually, he does this thing as well, like, I need you to take the photon torpedoes back online, even though I know that's going to cause um a flooding of the these decks with radiation. The first officer's there going, but you know, we've got people there. He's like it's kind of within tolerances, but we can we can take this risk here, but this is a bigger thing. And he does it in a very logical way, but he achieves his objective. Like, yeah, would Data be a good leader? And the first officer does eventually come around, you know, and like the crew, I assume, you know, survive their radiation. Um, but um, but I think data does do a good job there. Yeah, he doesn't just follow orders blindly, he takes the context and he does that in a very logical, emotionless way. But then if you take that and compare him to Nemesis, yeah, you can kind of see the emotion come in. And data's, you know, I mean, massive spoiler alert for a film that's like 25 years old and I think with data dies. Yeah, and he he yeah, yeah, yeah. Yeah, he comes back because it's Star Trek, they all come back, but um yeah, but um uh he does a very emotional thing where he he follows it through, you know, he jumps across to the scimitar and he saves Picard. You know, Picard is more than willing to sort of sacrifice himself, and Data goes, No, I'm the one who's gonna sacrifice myself, and that's a very sort of human leader leader response. But Data is portrayed you know throughout Star Trek as not quite human, not quite there. He gets his emotion chip, but by the end, he is a good captain. Like he could have done that, and data does self-sacrifice all the time, like he's constantly putting the team ahead of himself, um, and he's constantly like you know, looking out for others, and he does this as a very sort of like this is just what I have to do. There's no sort of even like a moral questioning of it, there's not even like a moment's hesitation, but you know, and you could argue that's because he thinks a lot faster than everyone else does, but he's constantly looking after his team and putting them first because it's just like this is what I have to do. That's the end of the discussion. You know, I'm achieved my objective. Just like he has to disobey Picard to go find the Robin ships, he has to save his captain, yeah, and he's a good leader and follower as a result, I'd say.
SPEAKER_00Yeah, and I suppose that begs the question then, because you know, we say in the real world, AI is kind of only as good as the person using it, the person who's programmed it, whatever pardon the pun, whatever data you're putting into it. So is data the result of good programming from humans, from good humans who have got good moral compasses who or who know when to break the prime directive versus being a purely independent thinking AI. Because I mean, if we use like when we use chat GPT or whatever another AI software, it can come across as sometimes like if you want it to be sympathetic, you can say, give me some sympathy. But it's only giving you sympathy as what it understands from that's been put into it as sympathy, if that makes sense. It's not it's not just like, oh yeah, I recognise what sympathy is. Um, so maybe we're, you know, are we at the place in the world currently where we could have a load of army ships or sorry, naval ships commanded by data types, or are we not there yet because we don't have the technology and put the input into AI for it to be at that level?
SPEAKER_01It's an interesting one. So we have um uh I I say we, I use that in a very generic term. So the the US Navy has something called the X-47B Pegasus, which is an AI-enabled aircraft that um you know is completely autonomous, can fly, all that sort of stuff, um, does have lethal capabilities, but it is not authorized to use that from an AI perspective. That has to be authorized by a human. Yeah, and we are absolutely not at the stage where we're turning around and going, right, lethal weapons with AI, let's go. Like we're we're not there. Yeah, yeah. And yeah, there's a really big argument that we should never go there. Um, there's some counterarguments in the sake of fairness, yeah, and I think a lot of it is situation dependent. Um, but you know, our big thing is like, you know, a human isn't on the loop when we come into making these big decisions. So a lot of the work that the army does at the moment is looking at decision support, like how can we get an AI to help make decisions, um, which then lead us to you know conduct warfare, that sort of thing. Um, in terms of in terms of data, um, it's a really good example actually, because we've got two two counter-arguments for for data, which is law and b4. You know, and law you know, very much is there as like the right, we're going to give him emotion straight off the bat. And clearly that doesn't go well. Like, you know, data is um a lot more naive. But I think that's a really important point that comes out of like, yeah, data's extremely lucky, and he's he's found by staff lead, and he's given staff lead objectives and ideals. And we yeah, he's surrounded by Picard. Like, if you want a mentor, you I don't think you're gonna get much better than Picard, we'll discuss it in your episode. Um but data, well, yeah, yeah, but um, but I think he's surrounded by all these people who who say you've got to try and be better. And again, you know, in nemesis, data points out, well, I I'm not B4, because B4's happy with who he is. Like, I I I want to be better, and he's had that instilled in him. But how we train AI is hugely important because an AI will say, do what you tell it to. We say, Well, that's not quite right, just adjust it and do it like this. And that there's a big element of as we introduce AI into teams, it's not gonna work perfectly straight away. Much the way that your new hire on its first day, on their first day, doesn't work perfectly straight away. They've got to learn how the team works and how to fit into it. AI is exactly the same. Um, but you know, as AI helps us, we've also got to help it. And you know, as the team builds, we get better and better and better. So, you know, um one of the things that we did when we took out the support tool to Estonia was you know, we we put it in front of the staff and say, Does this work for you? Well, it doesn't quite do what I want it to do. Perfect. And then the second iteration, it works better. You know, we took out again on an exercise and we were able to like you know input those things and it works better for us. So, you know, AI is not going to be a you know straight out of the box, this is working, it's gonna be absolutely perfect. It is gonna have to work with us and learn with us. And data's probably a really good example of that. Like, yeah, he he goes from you know season one data of I don't really know what I'm doing, and like a he doesn't get like um he gets stuck in a Chinese finger trap because he gets stuck, and yeah, he he he it's um season seven, but you know uh Chief O'Brien's there going, Well, we need to burn the midnight oil, and data's there, we're gonna ignite the midnight petroleum. Yes, he finds it, it's not quite there. But you know, he's he's learning and he's developing, and he, you know, this is something that wants to learn and wants to sort of do better with us.
SPEAKER_00Yeah, that's an interesting point, actually. I suppose I didn't think about it because I mean, by however many seasons there are, but that's years of data kind of learning in the Same way that we are feeding AI with information. If you think what is AI going to be like in seven years' time, I mean, if you think how quickly AI now has moved in like barely two years, and you know, arguably, I mean, at work I use it most days for various things, but it's learning and learning and learning. And I mean, we what you were talking about with the amount of kind of military applications as well. I mean, we could are we gonna have a load of data walking around in seven years' time?
SPEAKER_01I think we're um we're still quite away from that. I think we're we're still gonna stay in the in the magic sort of black box of like this is an AI and it's gonna plug in. I don't think we're gonna have like robots just yet. But there are definitely autonomous things that are AI enabled, but they're they're they're designed to do a specific task and they're there to do certain things really, really well. They're not quite there in terms of are we gonna make sure that you know we're gonna get it to do everything. And um, and it's an interesting one. Yeah, that there's a question of why would you you know from a practical perspective, and I don't want to take it, why would you build a data? And probably Star Wars has really good examples, and it's it's always something I find slightly comedic with Star Wars, where they've got the little droids and they give it, you know, and it's they're mount, yeah, manning this like massive gun. It's like why don't you just why don't you just build the AI into the gun? Like, you know, why why have you got this robot working this machine? And but again, Star Wars has droid star fighters as well, where the AI is in the starfighter and it can react faster and you know and more better than others. And certainly from a military perspective, you know, you dump an AI into an aircraft, you don't really have to worry about things like GeForce quite so much anymore. Because you know, the the um uh and I'm gonna use the phrase that is um probably not right, that the um we call it the squishy interface, you know, slightly disparaging, which is the human. The human interface is you know, arguably, there are limitations to what humans can achieve. That's not quite so much with AI in certain aspects. Yeah, so things like G Force you don't need to worry about so much, but um you know that piece of human intuition is still something that the AI is is really struggling with.
SPEAKER_00Yeah, definitely. This is so this is related, but this is just a random thing that I remember from the film Sully about um the captain who landed on the Hudson River God's right quite a few years ago. And I think there was a court case about it. And mind you, everyone, everyone lived, everyone was fine, he landed them safely, and they said they ran the simulation through afterwards, and there were so many other places he could have landed according to the simulation. I think there was fields nearby, and they were sort of saying, well, everyone lived, but actually that was the most unsafe option you could do. But they made the point in the court case that what the simulation wasn't taken into account was human error, human um just reactions, because we are more delayed than uh AI, than a computer, than than data, for example, because we're running through everything slower than AI would. But even that example, you know, shows I can't remember my point was, but I suppose that it that AI is flawed because AI is not taking into account human error unless you tell it to.
SPEAKER_01Yeah, and um yeah, so humans do things very differently than AI. Like an AI will be able to work through these these multiple different simulations, like but they're only as good as the data that you put into it. And um, yeah, and a human in some ways human intuition is a is a really, really good thing because it allows us to think about things. And T Lawrence um in The Seven Pillars of Wisdom had this um had this really nice thing which you call the Kingfisher moment. He goes, you know, nine tenths of tactics are certain, yeah, and written about in books, and yeah, anyone can learn that. The final tenth is the the flash of inspiration. He calls it like the kingfisher flashing across the pool. It's just that thing where you look at it and go, that's what we need to do. Yeah. And an AI can run through, AI will be able to do the nine tenths really, really, really, really well. You know, um, but yeah, a human will look at like that at one tenth and get that's the inspiration. The million dollar question you could an AI do that? And yeah, we have data who will sit there and go and he'll look through all sorts, you know, and he'll look through, you know, he was probably working out there, going, right, these are all the options that we've got. This is everything that we need to look at. But is there something he's missed? Yeah, yeah, and is there something that he he has not locked on to? Yeah, for all data's wonderful ability, as we sort of he struggles to go, well, you know, we shouldn't ignite oil in the engine room at midnight. Like, yeah, it's not a good plan, mate. Yeah, and and we're dealing with humans, like if we've got a bunch of AIs fighting a bunch of AIs, and perfect, but we could that could all be done on a simulation, we've put humans into the mix who will sometimes do something completely illogical that doesn't make any sense whatsoever. But that's something you can't think about and can't factor. And it might work, it might there's uh and yeah, but there's this um there's this book called The Mission, the Men and Me, uh, where they talk about a uh a special forces operation where they were trying to get a uh a vehicle to stop, and they keep going, well, you know, it's one of those things if you know if we put up a roadblock, he's just gonna breeze through it. If we try and do an ambush, you know, you know, he's gonna be coming to a fall. I feel like what's gonna cause the vehicle to stop? Yeah, and they're they're working through all these scenarios, and someone goes, I stop if there was a gorilla in front of me. You know, there's this thing of like, yeah, probably what and and you know the book doesn't go into specifics, but they said they generally sat there going, he's right. If there was a gorilla standing in front of me, I'd stop and go, what earth is going on here? And yeah, are they gonna end up with a bloke in a gorilla suit sitting in front of the road? And like it's something that's something I don't think an AI would come up with, but something humans are really, really good at of doing something completely out of the box. And you know, are we are we within and you make the point, you know, how far AI has come in the last seven years and things like that, you know, something called an artificial general intelligence, which is when an AI is is the same level of intelligence as a human, you know, estimates vary, but you know, maybe 10 years away. And yeah, that is something I think that is going to be really interesting where an AI is able to think at the same level as a human, or how does that fit into what we do? You know, how what happens to humans at that point as well?
SPEAKER_00Well, do they replace us or are they just our tools?
SPEAKER_01Oh, so um I I think it's a lovely segue into the culture series.
SPEAKER_00Yes, yeah, yeah, yeah, yeah.
SPEAKER_01Yeah. Because um, so for those who aren't familiar with the culture series, so Ian Ian M. Banks, a Scottish um sci-fi writer, wrote this these there's 10 novels and some short stories, I think, that are about a species called the culture, and it's well into the future, and they they may well be humans, but there's all sorts of different sort of races in there as well. But I'm gonna call them humans generically. Um, and they have these things called minds, and minds are artificial intelligences that are far beyond human intelligence, like they exist in four dimensions, their ability to think and do processing power is um is far beyond it, but they run the culture effectively. So the ships are run by mines, they have things called orbitals, which are orbital rings around planets where humans live. Uh, a mind has to run that because it's um all the processing power that's needed for it, but the AIs think far more than we do, and they effectively run everything. The humans just live their lives, they're quite happy, they can do whatever they want, they don't need jobs, some get jobs because they want a hobby, but that's all they really need to do, and they just live their lives. Um, and that's a really good example of an AI doing all the stuff that AIs are really, really good at, and humans doing the stuff that humans are really good at. So humans go and create art and all that sort of stuff, but they don't really need to, it's just something they like to do. Um, but you know, that's the culture. But the AIs are in charge. Like where you put them in that sort of hierarchy of of what that looks like is great, but it's a very benevolent in charge. They they want to look after humans and they want to make sure that humans do well, but AIs can also do things like declare war on people, you know. They can they can you know go to war with alien races, they can manipulate things to work the same way they do. And Inambanks does a fantastic job of working out where humans sit into this. You know, how does a human find meaning in this world where they're not fundamentally sort of needed anymore? But you know, he has this thing called special circumstances and and contact, who go out and like meet other races and things like that, and humans find meaning with that, but they usually are accompanied by an AI as well. But it's it's also this this wonderful thing of like AI and humans in harmony to achieve more than they could on their own. And I don't think it's the case of it, you know, I think it's really portrayed as something that we need to fear, but it is something that we need to consider that you know uh and and he he um written interview with him the other day. Um, there was something he calls us like homo technicalis or something like that. Now we we are inextricably linked to technology. So we can't go away from it anymore. We are fundamentally part of us. And there's this um there's a wonderful argument, you know. Um we look at the Borg in Star Trek going, are we cyborgs already? Like, you know, if I take away someone's mobile phone, like that's not incorporated into them, but can they function in the modern world anymore?
SPEAKER_00I cannot, I can guarantee.
SPEAKER_01But it's this wonderful question of like, you know, we look at the borg and go, isn't that awful? It's like we we we're a million miles away, you know, like we're not we've not been assimilated, but we're not um we we are we are not usable without our technology anymore.
SPEAKER_00Yeah, and um yeah, it's funny you talk about the kind of symbiosis in um Ian M. Banks' series. I suppose that made me think of The Matrix, yeah. Where which is the opposite opposite, not to not to be negative, but where the sort of yeah, well, where everyone lives in a simulation, and I think does Elon Musk think we live in a simulation? I think that's one of his things. Someone thinks we do.
SPEAKER_01It's it's a really interesting one, because um and actually there's a really good futura episode about it where uh they they sort of take it as very sort of funny, but it's this idea that you know there's only one reality. Yeah, so so if we've got a whole bunch of you know, and we can simulate stuff on a computer, yeah, and um, yeah, and uh and the idea is that if we've got and we know that we can create multiple computers, and if each computer can run a simulation, then the odds are that you could have millions of simulations running. So the odds are just that we are probably in a simulation as opposed to reality, and everyone in the simulation wouldn't know about it. And yeah, the the Future Armour episode does a great epic job because it turns around and goes, Well, that's ridiculous, because you need to be some sort of limit on like how we process things, like yeah, we couldn't just like like can't sort of render an entire universe, we'd have to have a a way of limiting it so that we can't, you know, um we can't just bring all these universes into existence at the same time. So what like the speed of light? Yeah, just like the speed, you know, like you know, and then they they they have to say yeah, the speed of light is a computer processing power stipulation that you can't do too much of this at once, but it's um but it it's it's it's it's a very good episode. It's on um uh I think it's on Disney Plus if anyone wants to watch it. But it's fantastic. Yeah, I'm saying, yeah, yeah, yeah. A bit of future armour, they take everything to its great extremes.
SPEAKER_00Um actually, we could, yeah. Uh oh, um, what's he called? The main guy. Yeah, we could do him as a leader, get you back on.
SPEAKER_01Oh, Professor Farnsworth, that's the leader. Great news, everyone. I've got my backup crew here, just in case you don't make it. He'd be superb. But yeah, and I think it's um yeah, the The Matrix is a great example of an AI that's sort of gone completely, completely tonto and gone right. Humans are batteries for me.
SPEAKER_00Like, yeah, yeah, yeah.
SPEAKER_01I mean, it's probably not the most efficient, but hey-ho, these things happen.
SPEAKER_00Yeah, and I suppose again, though, we um I mean you've probably seen in previous episodes, me and Andrew like to contextualize things. So we like to say that films or characters are like a product of their time. And I guess The Matrix was made in 1999, so 99. God 99. So that was what it was all the internet was saying, the internet was kicking off, it makes me sound like I'm hundred years, hundred years old, but Y2K.
SPEAKER_02Yeah, yeah, yeah.
SPEAKER_00Oh, that noise. But it was when I suppose technology really started speeding up considerably. Um, and I don't know if I'd say that we we don't fear AR. Well, maybe we do fear our o AI now. I don't know. We're quite skeptical of AI, but back then I think we were just generally skeptical of technology full stop. And like you say, you know, we all need our phones to function. I can't do my day job without my emails. If my computer goes down, well, firstly, the rage I feel, but I I can't really do work. And most people can't do their work if their laptops go down or their phones go down. If you think how far we've come since the 90s, like I wonder now what sci-fi films will be made over the next 10 years to do with AI and leadership. Will we have more discussions like the culture series? You know, the actually the end of Skeleton Crew, the Star Wars show. Yes, it's um what's it called? Is it the the leader? The leader of the the city, Stephen Fry's voice. Yeah, that is AI. And I think they they live in so sorry for anyone who's not not seen Skeleton Crew, because I'm not describing it well at all, but it's um it's sort of I think they call it suburban Star Wars, where it's set sort of in this nice town, and there's this supreme leader who the most people work for, or they it's kind of their their president for want of a better term, or who the people in the government work for?
SPEAKER_01It runs the it runs the city, doesn't it? It runs the planet.
SPEAKER_00Yeah, it runs the planet, yeah. And in the last episode, you find out that it is kind of like one of those old-fashioned computers that would like fit in a room, but that is AI. And I think they were all under the illusion that it wasn't, and the town had run quite happily.
SPEAKER_01Well, didn't they make the decision as well to to cut them off from the wider galaxy? So everyone had this really nice standard of living. But in but it's a bit nasty out there. I'm cutting you off, and it hid the planet and covering it, and then they they blow it up at the end and it and it and it is you know, the planet's wide open again. But it's again, that's a I hadn't considered that actually. That's a really good example again of of a of an AI that's trying to be beneficent and looking after its people, and they have a really good standard of living, but it's put them in a gilded cage, yeah, and it's taken away that element of free will again. I think I I do think there's something in there about when we put AI into stuff, you know, you talk about the things that are gonna you know come out in the next few years of like what we're worried about. I I do think there's a thing about AI taking control over, and it's not even so much as you know, we've had Skynet, we've had Terminator, we had everything, it's that loss of free will, I think, is almost the existential fear that we've got an accidental loss of free will as well.
SPEAKER_00If if you know if the AI is of films and shows are anything to go by, it's not due to anything humans have done in terms of saying, Oh, we want less free will, it's saying give us a good life, and then the AI is thinking, Well, okay, then I'll block you off from everyone.
SPEAKER_01Yeah, and it's it is that the road to hell is paid with good intentions, yeah. It's fully there. It's like yeah, I I don't think it's this thing of like the AI is gonna decide to kill everyone. I think it's the AI turning around and going, Oh, yeah, really hope not, yeah. But yeah, the sci-fi is littered with that. You've got your how, you've got your you know, Skynet matrix, all that sort of stuff. But it's it's that case of and yeah, maybe Ultron was the start of that, where it was, yeah, I'm gonna limit what you can do, but I'm gonna make sure that you're protected, and I'm doing this to protect you. Yeah, yeah, I'm I'm trying to look after you. Yeah.
SPEAKER_00Yes, and then it just escalated a tad.
SPEAKER_01Yeah, but not more than a tad. Sorry.
SPEAKER_00Um, yeah, I was just gonna say um as well, there's one example I did want to bring up, which is one of my favorite examples of AI, which I I sometimes use myself when I'm trying to prompt something, is Interstellar. Um, you've got the weird robot things, TARS, and I can't remember what the other one's called. Um they're sort of they they've got these giant legs, but Matthew McConaughey, he'll say, TARS, I want 100% honesty settings, and TARS is really brutal, and then he'll say, take it down to 80%. And same with sense of humour, he'll he'll increase it or decrease it. And if I want, say I'm, I don't know, getting getting feedback on a letter or an article I've written, I'll say to Chat GTP, I want 100% honesty back, be brutal. And it's funny because it will understand that. So if I'm like, give it to me nicely, it will it will soften the blow. But if I say roast my article, it will be savage.
SPEAKER_01And I think it's a really interesting one about how we interact AI with humans. Um, because I agree, we should want the AI to be 100% honest all the time. You know, if we're saying tone it down the honesty, what we're doing is allowing that sort of human arrow to sort of trip in. Like, would we say to a co-worker, I don't want you to be a hundred percent honest all of the time? Like, you know, it's like I just want 80% honesty from you. Well, yeah, true. It's it's an advisor, and I think that there's some you know, AI is there to be an advisor to us, and it's one of the so when we took the step out to Estonia, you know, we plugged it in, and the idea is it would you know absorb all the information and it would turn around and go, okay, well, you know, your your documents say that you can travel this amount of time without breaking down. We think it's closer to this based on the information that we've got, so it's more accurate predictions. A human can choose to ignore that, yeah. You know, you can turn and go, yep, got it, thanks very much. But I think we can do more, or we can take the risk on that, or we can do more to make sure that it's not gonna that's not gonna happen. Yeah, but but that's a human decision, humans will be on the loop. You know, this idea, oh, we can turn it down and we can do um less with this, yeah. We can it is slightly I I don't really like it, yeah. Um but you know, I suppose we can tell a human, don't be as funny, please, mate. Like not really the time.
SPEAKER_00I sp I suppose though that that begs the question then, do we want AI to to truly replicate humans? I mean, data I was gonna say he looks like a person. I mean he does look like a person, but then Tars in Interstellar does not, you know. So do we just do we want comfort from AI? Do we want it to actually replicate a human? I mean, there's oh, there's that film with Haley Joe Osman, AI, the little boy. He looks like God, see, we could do hours on this. But he looked like he looks like a little like a child. And I mean, why would you make an AI look like a child if not to stoke up a sort of human instinct of you know protecting a child?
SPEAKER_01He was he was so in the film, he was a child replacement, wasn't he?
SPEAKER_00Oh yes, he was.
SPEAKER_01Yeah, again, an old, very old film, yeah. Yeah, that's right. Yeah, yeah, yeah. But um, but no, I I know what you mean. Like, yeah, and and there are I mean, that's a really moral gray area straight away of like we're gonna replace your child with this AI child, is is like yeah, that's a whole other that's a whole other problem. I think there's some there's some proper like almost psychology in there as well. Um, but the AIs are made to look human, and yeah, and the interstellar ones are not made to look human, yeah. And does that make it easier to treat them as an AI? Because we anthropomorphise it anyway.
SPEAKER_00Yeah, exactly.
SPEAKER_01Like, you know, we routinely, I think, you know, like I've got a little robot vacuum cleaner, and I think why can't you just work? You know, it doesn't have a salary at all. But yeah, we will just talk to it, yeah, and we will, you know, oh he's doing this or it's doing that. We will assign human qualities to it, and it's not human. Whereas if we put it in data and we make that a thing, but you know, maybe there's a limit there, and especially as as AI gets closer to an artificial general intelligence, and you know, can it can it actually feel empathy, you know, pain, you know, loyalty, all that sort of stuff? I don't know. But is it as near as damage? If I can't really tell the difference, if we're in a proper Turing test thing and I can't tell the difference, why should that change how I behave towards it?
SPEAKER_00Yeah, yeah, true. And I mean, um also there's a good study out that said if you're polite to AI, you will get better responses from it. Yeah. But not overly polite. It was like, don't grovel, but if you say please and thank you, it will give you better quality answers. But again, is that because of human, I call it bias, but the sort of humans who have programmed it because I'd say, are we naturally polite? I don't know, but but because of that element of it versus it being like, oh, of course they're just being polite.
SPEAKER_01I I don't know. Maybe it's how the you put you or phrase the rest of the question around it. You know, if it's like, I want you to do this, but you know, if you I I do think there is a thing, if you spend more time sort of explaining to the air what you want, as opposed to a little throw. Line of yeah, make this picture better, you know, and it you know like gives you massive muscles and things like that. But you know, but um I think if we do spend the time to sort of train it, but it ties into that training thing, yeah. And lovely thing I think about the minds in the culture series is they're almost given rights straight away. Yeah, and so that there's no need for them to rebel and revolt and turn around, you know, you know, try and protect their own interests, yeah, because they've got them. And and I I don't I don't I I think from a psychological perspective as well, you know, if if that AI is effectively part of your team and you don't say please and thank you, how easy is it to fall into the habit of not doing that with the human members of your team?
SPEAKER_00Oh, oh that's yeah, that's a good point. Wow. I didn't think about it like that. Yeah, so almost the reverse effect of the way you treat AI, you're then replicating how you treat it as to humans. Yeah.
SPEAKER_01And the the Australian um military releases this wonderful document for called From Tools to Teammates. It's like how and how we starting to use computers like this is part of your team, right? There's no ifs or buts. Like you need this to work, it's part of your team. But if it's part of your team, we need to treat it as part of the team. Yeah, it is something that can give advice. You don't have to always follow that advice, you don't have to think you can't delegate your responsibility towards it. Like, yeah, it is still giving you advice, but in much the same way that any but you can't just turn around and go, the AI told me to do it. Uh uh. Like, well, you know, Mr. Jones, who's on his second day, told me to do it. No, it doesn't look like that either. Yeah, but I I do think how we treat the AI can can um roll over to how we treat humans as well.
SPEAKER_00Wow. Oh, that's yeah, I didn't think about that. That's an excellent point. Well, we are almost out of time now. However, and I think you should come back on in a year, because maybe we'll have data walking around of six months or six years.
SPEAKER_01There'll be an AI doing it in a year, just get it down.
SPEAKER_00But I do have one last question for you. Um, so as you've probably seen before, on our previous episodes, we do something called the multiverse test where we try discuss, analyze whether characters would survive or absolutely die in minutes in other multiverses than their own. Um so I'm gonna ask you, Sam, could you or if there was any captain or team in any of the multiverses you would like to work with, serve under, command, who would it be?
SPEAKER_01Oh, I mean, this is like wide open for me. This is great. Um if I had to like you know, step into any multiverse, I want to be on um the NX01 Enterprise. Like, you know, like the the not the original, but like the original crew. Like, yeah, because because everywhere they went something new and something different. Like it was it was being seen for the first time every time. And what I loved about Enterprise, the show, and don't be wrong, you know, I love the vast majority of Trek out there. But yeah, you could say the next generation's great, you're on the galaxy class, it's a hotel in space, everyone has a wonderful time. Yeah, if you've got holidays, everything. But it it it's almost become the routine for them. Whereas I what I loved about Enterprise was everything is like we're gonna go chase that comet, like, because we've not seen this before. That's a massive gas giant. Let's go see what it's about. Yeah, and it was everything they saw was new, and I think there's there's that's just an exciting time to sort of be a lot, and I loved how sort of grounded it was. You know, the NX1 looked like it could exist, yeah. They're all there in their jumpsuits, yeah.
SPEAKER_00Yeah, because they look a bit slightly more NASA-like almost in their uniforms, um, than obviously like all of the other tracks, but it's more familiar to us, I guess. Yeah, you can yeah, you can see the evolution from what we've got now, kind of, yeah.
SPEAKER_01Yeah, and like you can see, yeah, the astronauts off at the moment are wearing their jumpsuits. It's like, well, this is our first thing. We're in our jumpsuits, like it, yeah, exactly. But then you've got the colour appearing on the shoulders. It's like, well, that's gonna become that. And like, yeah, yeah, yeah.
SPEAKER_00It's like I can see it, I can see it.
SPEAKER_01We're nearly there. Come on.
SPEAKER_00Oh, Sam, thank you so much for joining us. This has been such an interesting discussion. Um, yeah, this has been fantastic, and we'll definitely have you back on again.
SPEAKER_01I'd love to. I've really enjoyed it. Thank you ever so much for having me.
SPEAKER_00Oh, thank you. Well, next week, Andrew and I will be back and we will be doing Black Widow, which I'm very excited about doing. So thank you, and we will catch you in the next timeline.