This is a download from BFM 89.9, the Business Station. BFM 89.9, the Business Station. My name is Rich Bradbury. I am sitting in for Mr Jeff Sando in Studio With Me, of course, is Culture Pop's very own Matt Amitych.
Hey, Rich, how are you doing today? So far, so good. Let's see how things go by the end of today's show. So, the end of...
There's lots of happy places for everyone. I don't believe that. I don't believe what we're just saying, and you're looking at me like with this really hypnotic stare that scares me a little bit. But anyway, the editorial meeting for today's Matt's Blaine was the longest in history.
For 48 hours, we locked ourselves in the underground bunker that Matt calls his happy place, debating how to move forward with stories like Cambridge Analytica and the autonomous car scares. In the end, the answer was obvious as it was simple. Killer robots. It's time to Matt's Blaine.
I'm going to take a while guess here. Have you been watching the movie Geostome? I have, but that's actually more of a coincidence than anything else. It's a good movie.
It's essentially just Gerard Butler reenacting Olympus has fallen, but, you know, in space. Yeah, have you noticed that Gerard Butler and Duke Ray Scott are kind of interchangeable lead actors? You know, you've got Butler, who's the beefy one, but they do seem to do the same mixture of rom-coms and blockbusters. So I tend to think of Duke Ray Scott as being the thinking woman's Gerard Butler, whereas Gerard Butler is more the kind of gutter thinking woman's Duke Ray Scott.
So now Matt's Blaine is obviously a movie review show. Now, with dangerously off topic Matt, some of our listeners might want to know what goes on in the happy place we run about? Well, I could tell you, but then I would have to take you to my happy place and cue you. Possibly quite slowly.
It's a nice music in the background, maybe order something in from Uber Eats, exactly. And then on other nights I just watch Netflix. There you are. Matt Amity, let's find the killer robot.
How are those pesky machines trying to kill us now? Weirdly enough, there is a kind of link between Geostorm and the story that our listeners may have spotted this week about China's Changong-1, the space station that is due to crashback. Back to Earth anytime now. Now, I think we first mentioned this story last year, probably around October time.
And somewhat ironically, Changong means heavenly palace in Mandarin. It's an eight-ton space station that China put in orbit back in 2011 as part of its plans to become a super power. It wasn't manned, but it was designed essentially as a test unit. They wanted to be sure that their rockets could dock, that live support systems would function over long periods.
But it is habitable. China has sent astronauts to the unit to do experiments in the lab there. But back in 2016, China admitted it had lost control of the station and its orbit has been decaying ever since. It's due to hit the ground as early as this week.
If it hits between the recording of this show and the time you listen to it, don't switch off. We will get to the killer robots, I promise. Promises and coincidences from it, neither of which can be trusted, of course. Do we know where the station is going to hit?
Or even which station is going to hit? No, you know when you watch disaster movies like Geostorm or The Day After Tomorrow 2012? They'll always say something like, an asteroid storm is going to hit Paris on Tuesday at 1pm. And for some reason, Paris is nearly always the first to be hit.
Then there's somewhere in India. And then you go to the UK's Houses of Parliament and Big Ben will go spectacularly. It's either destroyed by a super heated space rock or a screaming kaiju. Where and when things will land is kind of an nonsense.
You can do some modelling of the likely outcomes, but you can't really say exactly where things are going to hit. And though it sounds weird to say, it's probably harder to determine where something like the space station, which used to sit above our planet, is going to hit compared to a meteorite or an asteroid that has been speeding through space for half of eternity. The heavenly palace is spinning around our planet at roughly 27,000 kilometers an hour. That means it could pretty much hit any part of the planet when it finally decides to plunge to Earth.
Luckily, most of it will get burned up as it re-enters the atmosphere, but it is possible that part of the ship may burn through and plunge to the ground. The parts might be the size of pebbles or potentially much larger. You know, how much risk is there that it might hit something like me, a person, or indeed a building? We'll go back to when you were a kid and the times you threw a lit firework on the ground.
You know, it's... I'd never do that. Of course not. It's spinning, it's sparking, no one knows where it's going to fire, so everyone runs in different directions.
And as you said, it's a really stupid thing to do and kids shouldn't do it. You've actually got a lot more chance of being hit by that firework than being hit by the heavenly palace. Most of our planet is water. Most of the land is still largely devoid of people.
So statistically, it is more likely to hit where you aren't. You don't need to run away from this firework. The odds I've seen range from one in hundreds of billions to one in 21 trillion. The odds are actually considerably higher that you're going to end up in my happy place.
And stuff falling from space happens a lot more often than you think. Apparently, the space station is only the 50th largest man-made object to fall to Earth. The largest was Skylab, the NASA space station that weighed a massive 77 tons and came down to Earth over Western Australia back in 1979, scattering bits over a wide area. A really nice end to that story.
A county in Australia sent a littering ticket to NASA because some of the debris came down on a public highway. Brilliant. You'd have to do it. Of course, no.
Right, so has anyone ever been hit by a falling rocket? Well, ironically, there's a fairly good chance that our first encounter with an alien species will be when a Tesla roadster crashes into their spacecraft somewhere out there in the Milky Way. It probably should have one of those tinted window strips that caused use to have in the 1980s. Elon would be over the driver's seat and alien on the passenger side.
Can you imagine the insurance claim for that? Sorry, did you say your car hit a Class V intergalactic cruiser? On its way to Beetlejuice. What's the speed limit around there?
To go back to your question, the answer is possibly. According to a story from Wired.com, a lady called Lottie Williams may have been hit by part of one of NASA's Delta rockets as she walked through a park in Tulsa, Oklahoma, in January 1997. She wasn't hurt. A small piece of something hit her shoulder.
It was never definitively identified as coming from the rocket, but the time and location coincided with the rocket burning up in that part of the atmosphere. Alright, so that's your public service announcement done. Thanks. Does the story relate to anything bigger?
Where are these killer robots you're talking about? Well, in a sense, a big autonomous space station is a killer robot. And now I'm cheating a bit with the definition here. I'm fudging, yeah.
But we're interacting with an increasing amount of this self-propelled technology. And there is a Darwinian effect where we rub up against a new and unfamiliar species, in this case robots and autonomous machines that are going to be casualties. I assume that you're talking about the incident with Uber's self-driving killer car. Yeah, which we discussed last week in weeks.
Now, there are very real and tragic circumstances that occur when we try out new technology. And that, of course, was one of them. Statistically, this almost had to happen when you have a tunnel war of metal traveling at speed and it comes up against puny humans with their protective layers of muscle and bone. There really aren't any good outcomes.
Yeah, we're a bit squishy, aren't we? We're very, very squishy. So something like this is genuinely going to happen at some point. And let's be honest, the driver of any kind of car is imperfect.
People make mistakes. We cause thousands of fatalities on roads every year. And the machines we're trying to replace us with are still in their test phase. So inevitably, they're going to make some mistakes as well.
In this case, the disaster was twofold. Neither the car's operating system nor its failsafe human driver were able to react in time to save a life. The tragedy is that this happened to Elaine Hertzberg. So you think that this is the kind of story we're going to be hearing more of?
Yes, we're going to see a lot more of these stories. In terms of vehicles, the one in which automation seems to be working most successfully is actually in aviation. And why does that work? Well, because of their size and speed, planes have to keep a vertical distance of at least 300 metres apart.
If the planes are flying at the same height, they should keep an amazing distance of 15 nautical miles apart. So even when the autopilot is on, the human crew should still be fully engaged. And if a plane lands itself, normally it does so on a runway at an airport. That's an area that's noticeably devoid of children's play parts.
What aircraft don't do is fly through shopping districts and residential areas, no's to tail with the planes in front and behind them and inches away from flights coming in the opposite direction. So in a lot of ways, the plane has the much easier job building powerful machines that interact with humans in our spaces is very, very risky. Now, do you think that it was possibly too soon to be testing this kind of machinery on a public road? I'm really glad to say that I don't have to know the answer to that question.
Companies like Google have been playing with this idea for quite a number of years now. So this technology is not desperately new in that sense. Most of the major car manufacturers do seem to be in this game in one form or another. Uber has been doing it for less time than Google, but more time than other companies in the field.
Are any of them ready for public testing? We don't really know. Was Uber the odd one out? We don't really know.
How do Uber systems compare to the competitors? Again, we don't really know. Uber system may be the best or it may be the worst. The likelihood is that it's somewhere in between.
What I am fairly sure of is that this kind of accident had to occur at some point. Even if the systems have been tested for decades and they're computationally perfect, it doesn't matter how many failsafe in place. At some point, metal hits meet and meat becomes mints. We discussed this a little bit on Geek Squarts last week.
So I don't want to spend the whole show doing it. Things and robots are coming into ever closer proximity. So I wanted to have a look at some of the results of those interactions and have a think about what we can expect from the future. When you say have a think, you mean things are going to get dark.
You're going to get dark. You're going to get dark. That's all coming up after the record course. You are listening to Matt Slain here on BFM 89.9.
I'm Richard Brent, recently sitting in on behalf of Jeff Sandhu. We'll be right back after these messages. Building First World Mindset. BFM 89.9 BFM 89.9 and the Business Station.
It is Matt Slain. I'm starting a studio with Matt Slain himself, Mr. Matt Slain. Of course from Culture Pop.
That's right, we're Matt Slaining on this sunny and possibly fragrant. From Friday, Matt is talking about killer robots. Yeah, my favourite topic. Are robots a fantasy or a nightmare for you?
I'll just start off with that. We're heading into happy place territory again. One Matt's fantasy is another man's nightmare, as they say. Who says that?
Very often the recently departed. On the subject of fantasies, I will be talking about sex robots at the BFM Health and Living Live event on Saturday, April the 7th. I'll be on a panel with Dr. George and a bunch of other people who are a lot more qualified to talk about this stuff than I am.
But yes, we're talking about robots with a murderous nature rather than a seductive one today. One of the most common places for interactions between humans and robots, of course, is the workplace. As usual, white collar workers are not the ones who are most at risk. That joy is reserved for the people who do the heavy lifting, or at least for the people who used to do the heavy lifting.
Industrial jobs. Yeah, I mean we've also seen those 1960s and 1976 comms where human secretaries are replaced by robots, but in reality it's the world of manufacture, especially the factory floor, where you're most likely to come into contact with the machines that we most closely identify as robots. We've long relied on machines to make everything from food to clothing to roof insulation, pretty much everything. But by and large those machines were human operated, and humans did most of the fetching and carrying as well.
A lot of modern and high technology factories now look more like labs. They feature a combination of strange color blinds and weird hieroglyphics on the floor. They're not symbols declaring membership of some villainous super club. You know, you haven't wandered into spectre.
Those lines and symbols enable the automated workforce to find its way around as it does all of that heavy lifting, leaving the human operators to lurk in a control room somewhere, eating donuts and discussing their similarity to homosimpsum. Now, one of the things that most of these factories have in common, whatever they're making, is that humans are not allowed on the shop floor while the machines are in operation. Some factories have automated systems that switch off the entire production line if the human wanders into the machine area. You know, we've all seen the YouTube clips of a robotic arm that picks up the raw egg without breaking it.
Those are propaganda videos. You know, these machines are usually custom made. A robot arm might put, it might swivel around nearly 360 degrees at high speed, with seemingly no warning, holding half a car. The home has sit in their cubicles, running programs that keep everything in sequence.
But those programs do not factor in slow moving donut eaters wandering around between machines. I'm guessing then that factory robots have indeed killed people. Health and safety laws were not always what they are now. So yes, you would be correct in that assumption.
It seems likely that the first person killed by a robot in this kind of environment was a guy called Robert Williams. A robotic arm struck him on the head on a Ford production line in a plant in Michigan back in 1979. Now the judge later ruled that the factory had insufficient safety procedures in place, and that's what resulted in Mr. Williams' death, and his family was awarded 10 million in damages as a result.
Not a great result, they would prefer to have. Right. Not around. Exactly.
In 1981, a guy called Kenji Urada was killed in a Kawasaki plant in Japan. It was reported that the machine he was killed by was actually turned off, and Kenji stepped over at he rope that kept people out of its area and knocked the on switch of the machine as he did. So, and of course, the machine came to life. It pinned and crushed him to death before his colleagues could rescue him.
Now, I'm not going to list out all of these things. You can Google it if you want to see the impressive toll of, you know, robot science. Where else do we often come into contact with robots then? Well, oddly enough, in hospital, the place that you might expect to end up after an unfortunate incident with a robot.
This one actually came as a bit of a surprise to me because I've been quite a proponent of robotic doctors and surgeons over the last few years. I've been forwarding the position that the machines are more reliable than human doctors, so I might have to rethink that position. I was at the same kind of thought process as well. Well, it turns out that robots have played a part in the death of close to 150 people between 2008 and 2013.
So, some of the deaths are what we would term as negligence in human doctors, such as parts of the robot falling into the body cavity or it dropping materials into the cavity. Other issues were more technical, such as machines freezing or needing rebooting, video feeds or cameras malfunctioning. You know, of course, the success is outweigh the losses, but it goes to show that a lot of this technology is what future map would probably think of as being primitive. Alright, if we can't trust the doctors, can we trust the police or the armed forces, the robots there?
Sure, why not who better to stop things from killing people than the police and the military? If you didn't detect the sarcasm there, even in the best case scenario, no. Some robots are designed to kill, and the argument being that it saves lives on our side and ends lives on their side. Most countries with modern military forces now seem to have some kind of autonomous weapon systems.
They have aerial drones for reconnaissance and on occasion, covert assassination and bombing. You get automated firing systems for tanks and artillery and for naval guns. We have bomb disposal and detonation systems that are all run by robots. I mean, when you look at modern jet fighters, those things are simply too fast and complex for human pilots to fly without enormous help from the computers.
So you could probably describe them as being as close to cyborgs as we currently have in that the machine needs both the human and the robot in order to function. You know, there's a huge range of things that both preserve and take lives, and it isn't limited to national military cyber. Terrorist organizations are making use of commercial drones such as the ones you or I might buy to irritate our neighbors, and those can be used to transport small items or even as improvised bombs. I believe that most of those systems are still guided by human operators.
They're the ones who decide when to shoot the gun or when to drop the bomb. Largely although not exclusively, nine members of the South African Defence Force were killed by their own artillery piece during a training exercise back in 2007. Now that's 11 years ago for those of you who think this kind of robotic automation has only been around for the last five minutes. The machine had a semi-autonomous system that allowed it to target and shoot at aircraft, presumably because that makes it faster and more accurate than a human gunner.
Now what happened has been contested, but it seems likely that the gun, while on its lowest elevation, malfunctioned somehow and swiveled round while firing and killing the team that was operating it in the process. Other police, any better? So far we don't have any of these kind of robot type instances. Thanks, Lee.
But private security firms have been using robots to patrol moles and premises with fairly mixed results. A couple of people in the US have been caught trying to tip over or otherwise vandalise the robots, not their finest moment. If the bad guys can tip you over, then your good guy credentials are going to be a little bit weak, right? There is one recorded case where police officers put a robot on the offensive.
That was in 2016 in Texas. A gunman called Michael Johnson had killed five police officers and injured seven others, rather than risk more lives in a standoff with the shooter. A bomb disposal robot was fitted with a small amount of plastic explosive and it was driven towards the gunman whereupon it was detonated. Is this where we talk about the ethics of all of this?
Don't get me wrong. I can see why they have done it. But that doesn't make it right. It doesn't make it right.
No, I know. For today I'll leave the ethics part to other people and concentrate on what a society with these machines might look like. Oh great. We'll probably look at the ethics another day because in a lot of these cases the robot was doing what it was supposed to be doing.
In fact, during environments there may be a lapse in safety that causes injury or worse. A human might knowingly subvert the safety protocols and get hurt that way. When you look at the South African case with the autonomous cannon, it may have malfunctioned a bit, but mostly it did what it was supposed to do, which is to shoot at anything that moves. That notion that we have a friendly and unfriendly fire, the machine doesn't make that distinction.
Its purpose is much simpler. This is one of your trademark dystopias. Possibly, although I think my dystopia is probably the opposite of what most people are imagining. When we think about movies like The Matrix and The Terminator, one of the things that's so terrifying is that these machines are working together with a sense of purpose.
They have intent, there's a plan, there's an end goal. That end goal is usually to eradicate the human race and make the robots the planet's top predator. It's kind of a remake of wacky races for the genocidal. That's a scenario we understand.
It's still the good guys versus the bad guys where the humans are invariably the good guys. Which means that it plays out like any binary story. Sometimes the good guys win, sometimes the bad guys get the upper hand. There's an ongoing struggle, there are two sides, and you pick a side or a side is picked for you.
It's that hero's journey nonsense and you're right, we have turned into a movie show. In a lot of ways, I think the situation that we're heading for could be even more frightening. Because there's no real sense of who the good guys are. More because the machines just don't care.
When you look at the Terminator franchise in the Matrix, there's still some kind of central command. There's somebody that we feel we can reason with in the Matrix. An accommodation is reached with the computer system so that humans and machines can coexist. In Terminator, who knows?
Every time they reach a resolution, they jump it in time and reboot it in another direction. In both of these instances, we have an enemy and the robots behave very much like a traditional human enemy. In our world, the machines are actually implacable. They've been programmed from a task.
You can throw Asimov's three laws of robotics and doing no harm to people straight out of the window. We're building machines that can't be reasoned with and some of which have no other purpose than to kill. Yes, we can turn them off, but that's usually after the fact. You're not preventing harm, you're just preventing more harm.
It's like one of those recurring Star Trek plots. There's this planet whose inhabitants have wiped themselves out, but whose automated weapon systems continue to target and fire on any ship that enters their atmosphere. Precisely that, those are the systems that we're currently building. If anything, we might actually be better off with a spy net, a control system with intelligence and irrationality.
Something that has an awareness of itself, something that has some kind of responsibility. The really scary robots are the ones that have a purpose and no brain. If we go back to the case of the Uber car and Delaine Hertzberg, for the machine controlling the car, that was data. It files it as something not to do again because that's how we've programmed the machine.
Change the programming and it would use that same information to target every pedestrian it encounters in the same way that it now avoids them. It's not that the machine doesn't care. It's that it can't. It can only do what it's been told to do.
If we build a spy net, if a spy net decides to hate us, at least that's something we can understand. If we can't turn it off, at least that's something that we can fight. Killer robots. Killer robots.
Fun Friday. Matt's playing here on BFM 89.9, the Business Station. Thank you for listening to this podcast. To find more great interviews, go to bfm.my or find us on iTunes.
BFM 89.9, the Business Station.