Advent of Computing - Episode 94 - Robots, and the End of Humanity

Episode Date: October 30, 2022

Robots have always fascinated and horrified humanity in equal measure. The prospect of a synthetic lifeform is at times exciting, but can quickly turn south. Luckily we've never gotten that far... or have we? This episode we will look at a selection of early robots, from the Mechanical Turk to Elektro. All have one thing in common: they run off smoke and mirrors.   Selected Sources:   Robots of Westinghouse by Scott Shaut - Best source on Elektro and his friends   https://www.youtube.com/watch?v=T35A3g_GvSg - See Elektro in action   https://www.google.com/books/edition/Inanimate_Reason_Or_a_Circumstantial_Acc/mvVdAAAAcAAJ?hl=en&gbpv=1 - On the Mechanical Turk

I've always been a huge fan of Isaac Asimov. I know, it may come as a surprise that I'm a lover of science fiction. He was actually one of the big reasons that I became interested in the sciences in the first place. I was first exposed to his robot stories in middle school, and I've loved them ever since. There's something comforting about Asimov's robots. At least, in his early works. This is all thanks to the three laws of robotics. You may have heard of them.

The first law states that a robot may not injure a human being or, through inaction, allow a human being to come to harm. Law two states that a robot must obey orders given it by human beings except where such orders would conflict with the first law. And law three is that a robot must protect its own existence as long as such protection does not conflict with the first or second law. These Asimovian robots are friendly to humans. They exist as these superpowered tools that serve us. But there's something unsettling, something a little creepy beneath the surface here. You see, Asimov's robots aren't just big metal machines. By and large, they're humaniform or humanoid in some regard. In his

science fiction, they eventually become indistinguishable from humans. The key difference is that these robots are strictly better than us. They don't age, they don't die, if they get harmed they can be repaired, and they're programmed to be selfless. That sets up some horror for down the line, but this isn't the scared-out-of-your-pants kind of horror. What happens to humans if we're suddenly outmoded by a better lifeform? What if we made this better lifeform? Where still, these robots are programmed to serve us. This ends up being a big plot point later on in the Foundation series. As robots became more and more advanced, they started to interpret the three laws more broadly.

They deduced that it was their responsibility to protect not only humans, but humanity as a whole. So, we get this situation in Asimov's fiction where humanity is living under the benevolent rule of synthetic lifeforms. We effectively become a managed species. At that point, does humanity even matter? I really like Asimov's machines because they speak to a certain duality. A mix of wonder and existential dread. A mix of wonder and existential dread. These machines are designed to never turn on their creators,

but that leads to a possibly worse fate for humans. Stagnation, isolation, and control. I think that underlies the real unease that we have around robots. Sure, they're amazing tools, but what if something goes wrong? What if they become too human, more human than us? Despite this fear, there are many who have pushed robots towards fitting our own image. At least, that's what proponents will have you believe. I think under the surface we'll find there's a healthy dose of trickery going on. Welcome back to Advent of Computing. I'm your host, Sean Haas, and this is the conclusion to Spook Month, episode 94, Robots and the End of Humanity. In this episode, we're going to be examining a primal fear that's always lurking beneath the surface.

That's the fear of humanity's imminent demise and or replacement. I mean, it's all part of evolution. It's just a matter of time until it happens, right? But what does happen when something comes along that's better than us flesh folk? Maybe it's aliens, maybe it is just evolution, maybe we get lizard people that are just cooler and smarter than us by a little bit, or maybe it's an entity created by our very own hands. What's more ironic than humanity leading to its own demise. Humans have always had an obsession with this idea of manufactured life. This idea has gone by many different names throughout history.

Homunculus, Gollum, synthetic biology, but one we'll focus on today is the robot. Researchers and engineers have been working towards more human-like machines for centuries, but we still really haven't gotten there. I mean, Asimo's kind of cool and Boston Dynamics is doing some weird stuff, but we don't really have the slot-in human replacement yet. However, there have been numerous claims to the contrary. Most of those claims, however, have proven to be backed up by smoke and mirrors. In this episode, we're going to look at what I think is one of the most fascinating of these human-like robots.

Westinghouse's own Electro. This robot could do it all. It could talk, understand human speech, walk, and it even chain-smoked. Did I mention that Electro was built in the 1930s? Now, Electro himself is something of a fascination for me. I can't even remember how I first ran into this robot, but oh boy. Is this a strange machine? Electro was shown at a few World's Fairs, so we have wonderful recordings of this entity itself in action.

It's hokey and strange in all the right ways. At first, I kind of laughed Electro off as a dumb tech demo. However, there's more going on here. Also, did I mention Electro had a pet dog? going on here? Also, did I mention Electro had a pet dog? Anyway, Electro is part of a lineage of strange human-shaped machines built at Westinghouse. How did these machines come about? Why have us humans been pursuing robotics for so long? And why the underlying fear? I'm hoping to at least brush up against all these questions in this episode. Before we get started, it's the traditional announcement corner. As I have been saying for a few months now, there's still an open call for authors for

Notes on Computer History. That's the publication that I'm working on as a place for anyone to write about the history of computing. Notes on Computer History is intended to lower the barrier for entry for scholarship. I want anyone who wants to contribute to write in. That means you, dear listener, if there's something about computer history that you find fascinating or something you'd like to learn more about, then go do the research, write me a draft, and send it in. You can get all the information about where to submit at history.computer. With that said, let's get into the

episode proper. Robots are, at least in theory, pretty old. Folkloric accounts of mechanical entities go way back into the mists of history. The Iliad even includes automata, self-propelled machines that exhibit properties similar to lifeforms. Suffice to say, our species has been obsessed with robots since way back in the day. The only problem is that the marketing around these miraculous machines doesn't live up to the reality. I've said before that the key to making modern computers work is illusion. A single processor can't really run multiple programs at once, but some tricky software allows for the illusion of multitasking. This illusion's been bolstered by things like windowed programs, or networked file storage,

and other digital sleights of hand, maybe sleights of bits. A computer on its own just isn't very useful or even all that impressive, but we've been able to work around that to find ways to make computers truly useful and invaluable tools. So too, illusion has played a key role in the history of robotics, or at least in one branch of that history. Technology has never been able to keep pace with human imagination. It's probably a good thing it can't. The space formed there, between imagination and reality, can be used to pull some tricks. In 1770, Wolfgang von Kimplin would find a fabulous way to exploit that opening. His implementation of choice was a machine that's been called the automatic chess player,

or the automaton chess player. More commonly, it's known as the mechanical Turk. automaton chess player. More commonly, it's known as the Mechanical Turk. Von Kemplen was a Hungarian inventor, at least, that was the parlance used at the time. I think it's best to put him in the larger category of engineer, broadly speaking. During most of his early career, he was actually a civil servant, but engineering was always around somewhere in the back burner. The spark that led to his most famous invention came in 1769 while attending Empress Maria Theresa's court. At the time, there was an illusionist doing an act to entertain the Empress and her guests. After the show, von Kemplen told the Empress that he could give her a more surprising and exciting demonstration.

That is, if he had a little time to prepare. The Empress took him up on the offer. The next year, he would demonstrate the Mechanical Turk at court to raucous applause. The automaton's main purpose was to play chess. That's what it's most well known for, despite having more features. The initial demonstration was to set a pattern for all future showings of the machine. von Kimplin would bring an audience into a room where the automaton was already prepared. The machine itself is somewhat upsetting to behold.

It consists of a cabinet with the upper torso of a man attached to one side. This man, the mechanical Turk himself, is dressed out in traditional Ottoman garb. Here's another fun detail. The automaton uses its left hand to actually manipulate objects, but its right hand is used to hold a long pipe. The implication here, of course, is that the robot smokes. It's a nice little touch to, you know, humanize it. Now, according to von Kimplin, the humaniform section of the machine is mainly for show. Von Kimplin would demonstrate this by raising the Mechanical Turk's robes to expose a pretty sparse skeleton of gears and levers. The actual business is inside the lower cabinet. Accordingly, the next phase of the demo was to show off the cabinet's

contents. This cabinet itself was covered in drawers and doors to allow access to its inner workings. Those would be thrown open and von Kemplen would invite his audience to come up and inspect the machine. One detail that's often remarked upon is that you could see all the way through the cabinet. The doors were symmetrical in that sense. A door on the front had a corresponding door on the back. This was a neat way to show that the machine had no false back, that it was just as it appeared. There was also a neat bit of misdirection on von Kimplin's part. Once the crowd was satisfied that there were no tricks involved, the machine would

be wound up by a large crank. You know, since it's run by a big spring, that's the only source of power that would make sense. An audience member would be invited up to play a game of chess with the automaton. A crucial part of this phase was the assertion that von Kimplin had no control over the Mechanical Turk once a game began. He had to occasionally step in to crank up the machine, you know, it needed a little more juice, but that was all. Von Kimplin would actually step away while the game was in progress. However, there is a strange kind of connection between the inventor and the machine. The source that I'm using in this section is a set of letters that were republished in 1784 under the title, Inanimate Reason or a Circumstantial account of the astonishing piece of mechanism M. de Kimplin's

chess player. It's an interesting read that gives a very period analysis of the automaton. Just to add some fun texture, this is also a bit of an aggravating source to go through. Not only is it from the 18th century, it's also been translated from German. Some of the spelling isn't exactly standardized, which on its own is a little weird. The biggie is that it uses the long S, which, despite sometimes dropping into sources from this period, I have never seen before. Apparently, back in the day, the letter S looked more like a lowercase f without the cross. You end up with what looks, fittingly enough, like an S that's just been stretched out vertically. A long S, if you will.

Initially, this was just how the Latin alphabet worked. The 1700s were a transitionary period, where the short S started to see some use. In the case of these letters, the modern S is used only when a word ends in an S. Otherwise, the long S is used. So a word like chess looks more like chefs. That's not really in the dictionary. Anyway, here's where I should digress. In inanimate reason, it's explained that von Kemplen had this tiny wooden box with him whenever he demonstrated the Turk. The letters call it a case. During chess matches, von Kemplen would occasionally open the box's small lid and take a look inside.

The assumption was that this device allowed von Kemplen to monitor something going on inside his automaton. Inanimate Reason paints it in a slightly different light. It has been pretty generally supposed that this case is a totally detached piece, merely calculated to distract the attention. The inventor, notwithstanding, has given me the most positive assurances that it is so independently necessary that the automaton could not play without it. And he further says that whenever he discovers this secret, the world will be convinced of the truth of this assertation. Or, if you like, afertation. Not the best English. Anyway, since day one, there was this mysticism around the Turk. Von Kemplen never came out and explained

how his machine worked in any detail, just that it did work and you could observe it in action. The small case, or CAFE if you like, was a nice ruse to add to this mysticism. The overall effect of the setup was enough to intrigue onlookers. Some even accused the machine of being controlled by Satan or other demonic spirits. There is an account in these same letters of a woman having to leave the room in a tizzy because, you know, there's a machine, it's moving, and it's of the devil. But that's all the setup. You see, the Turk could actually beat humans at chess.

The machine could outmaneuver expert human players. Yet, even in this core feature, von Kimplin slipped in some misdirection. In general, these were normal chess matches. The mechanical Turk always played white, so he got the first move. It's a pretty good advantage, but not one that should matter against a skilled opponent. The atom-ton's metal arm would reach out, open its fingers, then pick up and physically move a piece on the board. It was playing just like a normal human.

Here's one of the neat misdirections. Von Kimplin would, at some point, explain that his machine could tell when an opponent made an illegal move. The machine didn't let anyone pull a fast one on it. If an opponent tried to, say, move a bishop sideways, the automaton would shake its head. It would then move the opponent's piece back to where it should have been and proceed to take its own turn. This skipped the opponent's turn. That's not a normal rule in chess, but it's a neat way to show that the Mechanical Turk was versed in how the game must be played. The skipped turn could even be played off like a limitation. The machine's just so rigid that it, you know, it starts taking its own

turn. It gets a little bit confused. There are still some issues to sort out, you know. That's the usual fare that people know of the Mechanical Turk, but it did more than just play games of chess. Another feature was its ability to solve a chess puzzle known as the Knight's Tour. This was, at least at the time, a somewhat tough puzzle to solve. The goal is to solve a chess puzzle known as the Knight's Tour. This was, at least at the time, a somewhat tough puzzle to solve. The goal is to take a knight at any position on the board and then move it step by step so that it lands only once on every square. You can't have the knight touch a square twice. Von Kemplen would invite audience members to place a knight anywhere on the board, and then the Mechanical Turk would proceed to solve the puzzle.

Someone would also place markers on each square the knight visited in order to ensure that the machine did, in fact, complete the tour. Not super flashy, but it's a neat trick that could further solidify the automaton's skills. The final act, and what I think is my favorite, is that the mechanical Turk could talk. At least, in a sense. After a rousing game of chess, von Kemplen would place a sheet of paper over the chessboard. On that sheet was a grid of words. The audience was then instructed to ask the automaton simple questions, to which the machine would respond by pointing to specific words. It's a neat feature that buttons up things nicely.

Not only can the machine beat a human at chess and solve arcane puzzles, it can also carry a simple conversation, truly a marvel of engineering. Or at least, it would have been. That is, if the Mechanical Turk wasn't all a ruse. It seems that the trickery here was suspected from the outset, or at least there were detractors. This skepticism even shows up in inanimate reason. Quote, "'Tis a deception, granted, but such an one as does honor to human nature. A deception more beautiful, more surprising, more astonishing than any to be met with,

in the different accounts of mathematical recreations." It's interesting to see this in the selection of letters because, in general, inanimate reason tends to be pretty agog about the automaton. The machine was publicly debunked as early as the 1830s, so von Kemplen and subsequent operators were able to keep the act up for quite a while. The initial suspicion was that there was someone inside the Mechanical Turk's cabinet who actually controlled the machine. At least, that seems to have been the usual snap reaction. This is why von Kemplen went to the trouble of opening all the doors and lifting the automaton's

cloak. He wasn't just trying to show you a little bit of robotic butt. He was attempting to show that there wasn't space for a hidden operator. The trick worked pretty well. An animate reason reported that the machine had almost no empty space inside it. Certainly no space for an avid chess player. Some guessed that maybe a child was secreted away under those robes and gears, but von Kimemplen was quick to show all the machine's inner workings. Surely, there just wasn't enough space. Here's how the trick was pulled off. Von Kemplen actually did stuff an entire living chess aficionado into the Mechanical Turk. There were two parts to this deception. The concealment of the operator and the control of the machine itself.

The trick of keeping the operator hidden simply came down to how the machine's cabinet was constructed. From the front, you can see a set of doors, each of which could be opened to expose the Mechanical Turk's innards. It's a little ambiguous as to how many doors there were, but there were at least two. What can I say? Ye olde engravings aren't the most solid sources. Of course, you don't want to expose all of the machine's guts. That would reveal the very real, very fleshy guts inside. So the human operator sat on a sliding bench.

In later engravings, this is shown as a little dandy man with a powdered wig sitting on a tiny velvet seat. It's all very fancy. The trick was that the operator could slide around and reposition themselves within the cabinet. This way, as von Kimplin was showing off the machine's workings, the true brains behind the operation could slide out of view. The illusion was helped along by some fake gears and pulleys that could be slid into place to further hide the operator. The controls of the Mechanical Turk were, I gotta say, pretty cool. The chessboard was built into the top of the cabinet, and it was normal in most respects,

except that it was built from a pretty thin material. This allowed a set of magnets to track when pieces were moved. These magnets, however, didn't connect up to anything mechanical. Instead, they gave the internal dandy information about the state of the game. They were able to see which pieces were moved where. Inside the cabinet was a tiny chessboard which the operator used to track the game that was unfolding. Along with this board was an armature that could be used to move the automaton's arm. It was set up such that moving the control arm to a piece on the internal board would carry out the same operation on the larger external arm. The machine also had a set of fancy-looking gears

and pulleys, but those were just part of the act. As the Mechanical Turk moved, a clockwork mechanism would be activated in order to make impressive-sounding clicks and clacks. It's all part of this larger illusion. The final wonderful bit of magic was a candle. You see, it was pretty dark inside the Mechanical Turk. Too dark to play chess by. So von Kemplen placed a small candle inside the cabinet, plus proper vent holes so air could flow. You know, you don't want to kill someone inside of your contraption. Inanimate Reason even comments on the candle, assuming that von Kimplin had placed it inside

the machine so that it was easier for spectators to see the Turk's gears. However, the light was actually needed so the machine's operator could see. Overall, I think the Mechanical Turk offers a look at how illusion can be used to maximize the impact of technology. Sure, the thing was mostly trickery. With the technology of the time, it simply wouldn't have been possible to make a robot that could beat a human at chess. We wouldn't get to that level of technology until well into the digital era. So von Kimplin fell back on illusion. The machine was designed in such a way to convince an audience that it was, you know, far too complicated to understand. But it had to be a marvel of engineering, right? Small details and weird mistakes were added to help the illusion.

Why does the mechanical Turk play with his left hand? Well, von Kemplen would explain that he intended the automaton to be right-handed, but had accidentally built the thing the wrong way around. Why did the automaton always play first and steal turns if he made bad moves? Well, it's an early model, and surely it won't hurt to give the poor machine a small advantage. Despite the tricks, despite the overall illusion, the automaton still had some neat bits of engineering. Here, I think it's best to look at illusion as a way to stretch out thin technology. Take the automaton's playing arm as one example.

Internally, it was a very complicated panograph, one of those levered toys that you can use to duplicate and enlarge an image by tracing the original. That's a simple device. You can actually make one with four sticks and a pin. But the basic idea can be expanded. The Mechanical Turk's pantograph was twisted and turned such that it transferred motion from a lower internal chessboard up to the larger main board above. It was also extended to allow for fully 3D motion.

Normally a pantograph is two-dimensional and only really transfers motion in one plane. Von Kemplen was just extending that idea. It also had the complication of a working hand. The automaton could grab and release chess pieces, so that's another complication to the machine. Sure, the Turk was mainly smoke and mirrors, but it was still a complex device. The final trick to the whole affair, and what I'd say is the most important part, is what the Mechanical Turk represented. It was a humanoid machine that could beat humans at our own game. Sure, it was just one game, but that's a type of encroachment that us territorial flesh folk simply cannot abide.

We have this weakness, you see. When we're emotional, we're more susceptible to the power of illusion. Let's now jump forward a number of years, leaving the mechanical Turk behind in the dust. We're heading into the 1920s to visit Roy Winsley, an engineer at Westinghouse. We're also picking up a new source. Most of this section, and the rest of this episode really, is only possible thanks to historian Scott Schout and his book Robots of Westinghouse. This is, as near as I can tell, the only text that seriously deals with Electro and the surrounding context at Westinghouse. But we can't jump directly to Electro himself.

We must start with an earlier machine, Winsley's so-called Televox. The Televox itself is a machine of utter beauty, and the kind of device that could only be the result of old-fashioned American engineering. Wensley was involved in research aimed at automation. This started with the automation of electrical substations. The idea was to replace humans doing menial tasks with simple devices. In the case of an electrical substation, that's an easy proposition. At the time, there wasn't much complication to the electrical grid, so substation workers just had to ensure nothing was burning down around them. Maybe they occasionally flipped a switch when a readout

dictated the proper, you know, event. This kind of work is a fantastic, low-hanging fruit when it comes to automation. This is the type of job that's important, it literally keeps the lights on, but it isn't satisfying as a career. It's also not hard to monitor readouts and flip switches, it's a simple but repetitive task. During the mid-twenties, Winsley would create automatic systems that replaced these human operators, which in turn passed the buck down the chain of command. Automation can actually lead to increased complexity, somewhat paradoxically. Cybernetics would one day come into the picture to provide a way to deal with this paradox, but we're not really there yet, and even if we were in the right time frame,

it's a different track. You see, once you automate subsystems, you get to the point where you need someone to monitor and have a little control over all these electrical switches. Let's take maintenance tasks, for instance. What if you need to shut off power to a section of the grid so an electrical pole can get replaced? If you had a human operator, then the central office could call up that person and say, Hey, turn it off. But these early automated systems couldn't handle that. They just sat there doing their repetitive task with no input.

This led to the birth of the Televox, an improved automated circuit that could be controlled via a phone line. The Televox, at least in this early incarnation, was a remarkably simple device. We're talking a box full of relays and some passive components that just sits between a phone line and a power jack. It worked like this. An operator would place a call to a phone connected up to the Televox. This call went over normal phone lines and passed through a central office like any other call. When the Televox's phone rang, the machine would physically lift the receiver off the hook and respond with three short beeps. That was a signal that the Televox was ready and waiting for a command. The operator would then send their command down the line as a

series of beeps in response. A circuit inside the Televox filtered out everything except a specific frequency, a control tone. The operating principle here is really similar to how early automated phone systems functioned. The Televox would count the number of these tones. Every time a tone at the right frequency came in, it advanced a gear. There was a small handshake, the Televox would repeat back the number of tones it heard and wait for a confirmation tone, then the proper action was taken. In other words, you just blast a number of tones that corresponds to some action, then you tell the Televox to execute that by blasting a final do-it tone. Once everything was said and done, the Televox would close a relay on its end, thus doing, well, something. That's kind of the neat part about the Televox. It was never a very specific device. It was built as some smart glue to hold together other machines.

In the substation example, you might have two actions configured. A single tone would turn on power to the grid, and two tones would turn that power off. You could just as easily rewire things so that a single tone would turn on some lights and two tones might turn them off. It all comes down to how the Televox was connected up to the outside world. Now, this does lead me to an interesting parallel. The first automatic telephone exchanges. lead me to an interesting parallel. The first automatic telephone exchanges. Back in the late 1800s, early machines showed up that could automate phone switchboards. These machines worked on a very simple dialing protocol. To dial a number, the phone's circuit would be opened and closed a number of times corresponding to, well, the number you want it to dial. This wasn't an auditory signal, but it's the same kind

of series of pulses that Winsley was using for his Televox. Automatic switchboards wouldn't be adopted for a number of decades. Winsley even describes how a Televox had to be connected to via a human operator. I just want to point out this weird connection. And, well, the parallels don't stop there. The Televox was tone-operated. In practice, this meant that it was controlled by playing a whistle into your phone. The machine would respond with its own set of beeps and boops. We even have these great publicity pictures of Winsley blowing an assortment of little tin whistles into a phone. It's all very reminiscent of freaking, the latter developed pastime of breaking into frequency signal telephone systems. It's even been reported that Winsley eventually

learned to signal the Televox by whistling down the line. So hey, maybe we have an earlier incarnation of the whistler on our hands. The connection here is superficial, of course. It's neat, but not of real substance. However, we do have something a little more interesting going on here. In a 1928 lecture, Winsley described this interface as a language. To quote, We couldn't build a machine to understand language, so we had to create a new language. Sort of an electrical Esperanto, as it were.

A sort of universal language. Pure musical notes are, of course, the easiest to pick out, and therefore the language we used in speaking to this machine, a composition of pure musical notes. End quote. This, dear listener, really rustled my jimmies. Yes, of course, there is the neat invocation of illusion here. Wensley was looking for a way he could talk to a machine and ended up flipping the problem around to create a novel solution. That, given a little coat of paint, you can call a language. But what's of interest to me is how Winsley is discussing a new language used to communicate with machines.

You see, there's an often overlooked passage in As We May Think. This article is most well-known for its description of the Mimics, Vannevar Bush's futuristic machine that stored data as a series of linked pages. That's cited as an early contribution to the idea of hypertext. The Mimics is very important, but it's not the only prediction for the future that Bush makes in this article. There's a section where Bush is discussing a machine called the voter, a device that can imitate human speech. He muses about the possibility of going the other way

around, creating a device that can understand the spoken word. Then he veers off into this. a device that can understand the spoken word. Then he veers off into this. Our present languages are not especially adapted to this sort of mechanization. It is true. It is strange that the inventors of universal languages have not seized upon the idea of producing one which better fitted the technique for transmitting and recording speech. better fitted the technique for transmitting and recording speech. Mechanization may yet force the issue, especially in the scientific field, whereupon scientific jargon would become still less intelligible to the layman. End quote. The attempt at universal languages include, in another neat connection, Esperanto. But that's neither here nor there, really. Bush is basically saying that we may need to develop a spoken language that's more fitted to automation,

a human tongue specifically for talking to machines. Crucially, as you may think, is published in 1945, well after the Televox hit the scene. You may see why my wheel started to spin on this one. As far as I know, there isn't a direct connection between Bush and Winsley. I can say that with some confidence because I have a secret source stash. Later in life, Vannevar Bush compiled an autobiography called Pieces of the Action. The book was prepared from an oral history series that Bush participated in, assumedly at MIT. At some point, the series of interviews was transcribed and filed into MIT's archives.

I happened to have a scan of those interviews, which are basically a more raw, extended version of Bush's autobiography. Whenever I run into a possible Bush connection, I crack open those files and do a quick search. You know, on the off chance that something turns up. I can't find any Televox connections in these papers, so I doubt there is one. That said, this points to something more interesting for me. The idea of a machine-specific spoken language wasn't some one-off. It seems that this

was floating around in this early era of automation. I'm sure there's more to this, but that would take us on yet another tangent that I'd like to avoid for now. Let's just say I'm going to keep an eye out for more mentions of human-machine languages. Getting back to the topic at hand, here's why the Televox ends up mattering. As I said, it was a very flexible device. Since it just sat between a phone line and some wires, it could be set up to do darn near anything. But this is a double-edged sword. Wensley had this really useful device, but it's not very flashy. That is, you'd assume that it wouldn't catch any public attention. But that assumption is very wrong. You see, the Televox

was replacing human operators. In some cases, this would have been seen as a bad thing, a story of how technology was driving people out of jobs. In this case, those jobs sucked. No one wanted them. So Televox was more seen as a useful tool that could replace humans. Maybe you can see where I'm heading with this. Talk of the Televox started spreading around Westinghouse as Winsley gave internal demonstrations. Eventually, a company publicist heard about the machine. And that's when it received an interesting moniker. The Mechanical Man.

I mean, it's mostly true. The Televox was able to replace a human through electromechanical means. It could only do this trick in a very limited sense and for a very limited set of tasks, but hey, those are just details. This is what made it into the papers. Here's a taste from a 1927 article in the Brooklyn Daily Eagle. Quote, Correctly tuned peeps and buzzes will make mechanical man perform 22, 200 acts. He'll serve a pre-Volstead drink, clock junior's bedtime, or the weather in mid-Atlantic for ocean flyers. Radio apparatus invented by Roy J. Winsley converts tones into actions, or is conveyed to it by any telephone, distance no barrier to use, but money might be. Car serviceman, for instance, would cost $40,000 to install in Flatbush Home. End quote.

Once again, none of that is technically wrong. The Televox was just some automation glue that could bring together any number of machines. It could also do automated tasks in general. You could use a Televox to remotely mix a cocktail, for instance. But you'd need to first build a machine that could do the actual mixing of a pre-Volstead drink. Wednesday's device wasn't exactly the most intelligent thing in the world, but it was enough to act as a control mechanism for pre-existing electrical devices. Put that way, it's pretty boring, so the company's publicists just said it was a mechanical man.

That has a nicer ring to it without being totally incorrect. Now, Winsley was a much more smooth operator than I will ever be. If I was in this position, I think I'd knock over some of my co-workers' desks. I'd at least shout at them a good deal. We're approaching territory where the actual machine can't live up to the hype. Instead of backing down or trying to handle the perpetrators, Winsley decided to just kind of go with the whole mechanical man thing. This was helped along by a little more corporate pressure. The higher-ups at Westinghouse wanted Winsley to give a public demonstration of the Televox, specifically at

the opening of a new club in New York City. These same execs also didn't want to throw Winsley extra funding to prepare for said demo. A plan had to be hashed to somehow make this actually very boring technology a little more spicy. So Winsley decided to lean into the sensationalized media coverage around his invention. If the press said he had built a mechanical man, then sure, why not show them a mechanical man? The final piece of inspiration came from a comic in the New York Times. One of the freakout articles showed a depiction of the Televox as a very physical mechanical man. We're talking a straight-up robot. The Televox itself, just a big box full of tubes, made up the body. A square head and blocky limbs were added to complete the look. Winsley got probably a sensible chuckle out of

the whole thing, so he decided to just lift the entire idea. The final demo would be the birth of the Westinghouse robots, quoting from Winsley. As it was only symbolism, I made no attempt to have the figure resemble an actual being. It was but a piece of wallboard crudely painted in loosely dangling arms of the same material. The opening was to be on Washington's birthday, therefore it would be well to tie in the illustrious founder in some manner. And remembering the Broadway witticism so well exemplified by the peerless George M. Cohen, the flag is saved, many a show. I arranged a simple mechanism with a counterweight in line so that at my whistle of a command,

the arm of the figure would unveil a large portrait of George Washington. End quote. So here's the picture I want you to have in your head. The Televox is just a box of circuits held up on a table. Lindsay had made a plywood cutout of a 1920s robot to put in front of it. Just this big, blocky, silly-looking thing. The robot has a big hole in its chest so you can see the Televox inside. Winsley had also jerry-rigged a weight pulley and probably a little motor so that one of the robot's arms can move.

At a whistle, which was probably just done with his unaided mouth, Winsley would throw the televox into motion. The arm would move, a flag was raised, and in doing so, exposed a painting of George Washington himself. This is some really weird and hokey stuff, but hey, it works. The crowd lost their collective mind. The media went into a frenzy. The age of the mechanical man was here. Robots were coming, and they were going to revolutionize how we lived. Best of all, these robots were red-blooded American patriots. Well, they would be if they had any blood. The robot also got a name, Herbert Televox. Now, we could write this all off as some weird craze. Stranger things have happened. But

let's look at the context here. Mechanical automation wasn't really a thing at the time. We're talking 1928. Cars are still made by hand. In fact, the assembly line was just over a decade old at this point. Phone lines were still, by and large, routed by human hands. 1928 is the same year that sliced bread is invented, so these are early days. So, these are early days. There was a lot of hype to go around, but the Televox was a really neat machine. There were soon hundreds of installations. The most complex one I've read about was a water district in Washington, D.C. Their Televox network could actually read off water levels of various gauges over the phone, all in beeps and bloops. Wensley's time with Herbert Televox, though, was relatively short-lived. The inventor would be drawn off to other work at Westinghouse, and he eventually left the company.

Herbert, his precious son, set a trend that would endure long after Wensley had moved on. Westinghouse was now a robot company. At least, they would keep robots around for certain purposes. The final thing I want to point out about Herbert Televox is this. He smoked. Robots of Westinghouse includes this lovely photo of Herbert smoking a cigar while wearing a fedora. Truly a sight to behold. And it establishes the robot's bona fides. Seriously, this is a pattern we have to keep in mind. All robots have to smoke. There are a few transitionary robots that hit the scene after Herbert, but we don't get a real successor until 1931. That successor was named Willie Vocalite,

and he brings us into a new era of robotics technology. And yes, Willie was a true robot, because he smoked cigarettes. This is the point where I kind of hit the high strangeness level that I was looking for. I'll quote directly from an ad run for a demo of Willie. Quote, He's coming Tuesday. He's not alive, but he talks. He sings. He moves. He smokes. He has no soul, no heart, no brain. He's a scientific marvel. Don't miss him. End quote. This clipping is, of course, complete with a sketch of this tin can looking dude puffing on a cigarette. Welcome through the looking glass, I guess.

All jokes aside, Willy Vocalite is a pretty fascinating piece of technology. This robot was designed by Joseph Barnett, who had previously worked with Winsley, so there is a lineage going on here. The core of Willy is an improved version of the Televox. Perhaps it's better to think about it as the next generation of the device. The beeps of the old units were gone. You see, Willy could actually carry on a conversation. He could speak, and he could listen. Or at least, it seemed like he could. Interestingly enough, von Kemplen gave up working on the Mechanical Turk to pursue a device that would emulate human speech.

The idea of a talking machine has always been exciting, that's the best way I can think to put it, and it's always just been part of this overall robotic milieu. Von Kemplen was attempting to recreate a human vocal tract. He built these bellows and reed-based machines that had flexible tubes. An operator could play these like a musical instrument and kind of mimic the sound of a human voice. It's this spontaneity, though, that's crucial. A recording device was one thing. a phonograph was certainly neat in the day, but it was a very static sort of medium. A machine that could produce human speech on its own, without a human in the mix, that would be its own kind of marvel. Von Kemplen had taken one

path towards that end goal. The nerds at Westinghouse would take another. Willy Vocalite could talk, but it was accomplished using a trick. The robot's speech was produced using a specialized 78 RPM record player. The record was pre-recorded with responses to queries, cool phrases and sounds to play along with actions. Tracks on the record could be selected automatically. If Willie needed to say hello, then the Televox inside him would just tell the record player to skip to the proper place, and a pre-recorded greeting would sound over a speaker. The pre-recorded message trick was initially developed to improve the old Televox. These ladder model machines were able to originate calls, but this was actually pretty hard to

automate. In the 20s, you couldn't just dial a phone number. You had to pick up the hook and talk to an operator. It would be nice if a televaux could call up someone if it encountered a problem, but it couldn't get past an operator by simply beeping at them. Thus, pre-recorded messages were added into the mix. This was initially done using repurposed video film of all things. This is a bit of a tangent, but I love the ingenuity here. Magnetic tape wasn't around yet, and vinyl had to be pressed, which was an annoying process.

Video film ended up being easier to work with. Back in the days of talkies, one option was to put an audio track physically next to each frame of the film. This audio was recorded on a free space on the film, so it had to be encoded as a series of light and dark patches. This was more convenient because you could create a recording using lights wired up to a microphone. The audio was then read using a photocell. It's cheap, easy, and simple to do. These tapes were recorded with a simple message that said something to the effect of, this is Televox at 1234 calling central office at 567. When an issue came up, the Televox would pick up its phone and start blasting that message. A telephone operator would eventually hear it and route the call. This would get the

Televox in contact with someone who could actually operate the thing. From there, they would drop down to the beep-boop language. The hacky part here, the weird tape stuff, is indicative of work done on the edge of what's possible. Engineers at Westinghouse didn't have a good way to make cheap customized recordings, so they had to adapt some existing technology. The hack was part of a more ambitious project. This tape-based voice would be incorporated into some of the transitionary robots, but Willy would really up the ante. That's part one of the illusion. Part two is tricky speech recognition. Willy was smart enough to recognize speech. A user could tell Willy to say hello, and the robot would respond back in recognizable English.

The first demos of speech recognition hit the scene in the 1950s, with the first practical system created some decades later. So how was Willy able to accomplish this feat before anyone else? Well, he wasn't really recognizing speech so much as patterns. I guess this means I should address the actual tin can in the room. I haven't really explained Willie the Robot so much as some of his guts. Willie looked like a literal tin man. He was made up of these stovepipe-like chunks of steel.

We're talking cylindrical arms, cylindrical legs, and a cylindrical body with a little cylindrical head on top. The poor dude had no muscle definition to him. The whole thing was riveted together, and he even had Westinghouse painted across his chest. The head is the part that I want to focus on. You see, Willie had eyes. At least, kinda. You see, Willy had eyes. At least, kinda. His eyes were photocells, simple devices that could detect incoming light. This was used to send data to the humanoid part of the machine. Willy's tin can body actually housed a limited number of components. Motors, gears, and some vacuum tubes, and the all-important photocells. He did have a Televox-like device inside his chest, but that was only half of his smarts. The rest was housed in a box that was

kept offstage during demos. This external unit was what handled speech recognition. The trick here, the piece that made everything work, was that speech was broken down into patterns of long and short pulses. The recognition device didn't actually know what a phrase like say hello meant, it just heard a short word and a long word, so it translated that into a short pulse of light followed by a long pulse. It's kind of like a simplified Morse code. Willie's electronic eyes picked up these flashes, and from there, the Televox setup took over. The long and short flashes are just easier to interpret with simple circuits, the kind that would fit into Willy's chest. Those flashes could instruct the robot

to play a certain part of the record, and thus could carry on stilted conversation. part of the record and thus could carry on stilted conversation. So there's the talking robot part that still relies on something like the beeps and boops of the older Televox unit. Willie just happens to use a different type of monotone language. However, Mr. Vocalite could do more than just talk. He had a small handful of tricks, but hey, it's something. Willie's legs were partially articulated, which allowed him to stand up and sit down. He couldn't walk, but he could use a chair. That's kind of cool. His arms were also articulated in a similar manner, able to move up and down slightly. It was also possible to plug other devices into Willie, but he could only turn these external machines on and off. So you could, in theory, have Willie control a vacuum, he just

couldn't push it around for you. The final interesting aspect is Willie's ability to smoke. Believe it or not, this feature was an advancement when compared to the older Herbert Televox. The one photo we have of Herbert smoking just shows a cigar held up to his painted face. Kind of funny, but far from a feature, not very fully fledged. We don't have much detail about how Willie's smoking mechanism worked, but he was able to actually puff. Pictures show that he had holes in his painted-on lips where he could hold a cigarette, and he at least blew some air through it. We have this one kind of strange photo of Willie releasing a plume of smoke near a child, so

there is evidence of this feature in action. The final thing I want to note about Willie is his bellicose nature. It's perhaps a character flaw. You see, Willie was packing heat. This detail is attested to by contemporary newspaper articles. Apparently, Willie could be instructed to fire a pistol. That's all the detail we get, just that he could shoot a gun among a list of other actions. Concerningly, it doesn't say if he could aim it or not. Taken as a package, we're dealing with an interesting machine. The illusion is easy to see through, but it still has an effect.

I keep finding myself referring to Willie as if he were a person, as if he could care. Willy fills the same space as the mechanical Turk. He was able to do things that humans could do and, in theory, maybe do them better. We don't have accounts of Willy beating people at games or anything like that, but there is the implication that this mechanical man could be a tireless worker. Willy was really presented as the next generation of home appliance, a machine that was posed to change our way of life. But that dream couldn't last. There's one other way that Willy imitated humans. He died. This is skipping forward in time a bit, which I think is important to establish the next part of our story, so forgive me for the discontinuity.

In 1943, Willie was scrapped as part of the US war effort. It's very possible that parts of Willie were used on planes that bombed Germany, or ships that fought over Midway Island. In that sense, he was a true killer robot. The reason for the sacrifice was simple. Willy was made out of steel, which was a valuable wartime commodity. That design choice sealed his fate. However, not all Westinghouse's robots faced a furnace. There is one that has survived into the modern day. That's the ultimate example of Westinghouse's work, Electro. Now, Electro brings us to a weird sourcing issue. I know, it wouldn't

be an episode of Advent of Computing without me complaining about sources. We don't have many written sources about the construction of Electro, but we do have these wonderful archival videos of him operating. We also have great photos of the robot and his innards. Conversely, we have good details on the history of Willy Vocalite, but no surviving recordings of the older robot in action. Most of the coverage of Electro focuses on his career, so that's kind of what we have to look at. However, we still know a great deal about how Electro works.

His general operating principle was similar to his predecessors. His heart and soul was a Televox-like device, backed up by an external control unit. He talked via vinyl and understood spoken comments by interpreting phrase length. He could also smoke, as is robotic tradition. Electro had a small hole in his upper lip for holding cigarettes, which were smoked via an air pump. The pump sucked on the cigarette and then blew smoke out of Electro's nostrils. My guess is Willie smoked in a very similar fashion. The biggest changes between Electro and his predecessor come down to mobility. First of all, Electro could walk. Kinda. There's always caveats with these dudes.

He moved via a pretty smart mechanism, though. Electro's feet each had a set of powered rollers. His right leg was basically just a fixed pipe. It couldn't move at all. His left leg, however, could bend and extend. This allowed Electro to scoot around with this awkward gliding gait. His left leg would move forward, and then he could drag his right leg up to match its position. Maybe not the most sophisticated thing in the world, but it's a neat way to use limited

technology to make a bipedal robot. The main issue they were overcoming was one of balance. Our humanoid bipedal gate is actually pretty complex. Try thinking about all the actions you take next time you take a step. As you prepare to bring your foot forward, it bends along the arch. Then you push off with the pad of your foot and its toes. As one foot is in midair, you have to adjust your weight to balance on your grounded foot. Then your lifted heel comes down followed by the rest of your foot. The process continues with your next leg. Then, your lifted heel comes down, followed by the rest of your foot. The process continues with your next leg. Turning is even more complicated. Westinghouse got around all of this by making Electro basically move on wheels.

The fact was hidden by rigging up his legs to bend. He sort of glides forward, which, without close inspection, looks like he's walking. I kind of love it. It's a little eerie to see him gliding around, but at the same time, it has this endearingly awkward quality to it. Some photos of Electro also show him wearing delightful little spurs, which I think were supposed to help stabilize early versions of his weird walk. Electro's hands were also more fully realized. He had rotating wrists and five articulating fingers on each hand. Now, he couldn't exactly grab things, but he could bend each finger at one joint. During demos, Electro would count on his fingers,

but in theory, he could have probably operated buttons and maybe typed on an extra large typewriter keyboard. The robot's head was also more expressive. His mouth moved up and down as he talked, and his head could rotate and nod. The general vibe here is that Electro just had more moving parts. He's a more fully actualized robot, a better mechanical man. He also looked a lot better than earlier machines. Electro's body was built out of aluminum, which gave him this characteristic shine. Instead of being a big tin can, he actually was molded to look sort of like a person. His legs and feet have some definition. His shoulders draw down to wrists and hands.

He has a big barrel chest. His head isn't just a painted pipe, but actually shaped somewhat like a human head. He's also massive. The dude is 7 feet tall and weighs 260 pounds. The overall construction gives us this weirdly endearing robot. He's this giant and lumbering art deco-ish aluminum oaf. I think this contributed a lot to Electro's appeal, and it's a strategy that was in use since the Herbert Televox days. Westinghouse's robots all have this clunky and friendly appeal

to them. They aren't here to dominate humanity, they're here to help. We can also contrast this to a more short-lived machine built in England, Eric the Robot, or maybe I should call him Eric the Evil Machine. Eric was built a little bit before Willie Vocalite, but you see, Eric had a branding problem. His creator, Captain W.H. Richards, was inspired by earlier Westinghouse-style machines, but he missed the friendly design note. Eric looked plain scary. He had a slack mouth with a few pointed metal teeth which sparked during demos. He also had a spiked collar and wore a simple metal crown. According to Robots

of Westinghouse, Eric was supposed to look like a knight in armor, but, quote, his expression on his face repulsed and frightened many in the audience. Not the best way to present new technology. I mean, even Willy Vocalite had a certain charm to him. The robot even had a painted-on smile. Sure, he liked to shoot things, but he was probably a gentleman about it, you know? It's clear that Electro's design followed a similar principle of being non-threatening. I haven't seen anything written about crowds being scared of Electro. I haven't seen any coverage about people being pushed out by this robot. Rather, most articles frame Electro as a modern marvel,

a machine that can do some cool tricks and something to help you around the home. He is a super-powered appliance, a portent of things to come, not a killer robot. Another factor that made Electro such a hit was his voice, or rather, what his creators recorded for him to say. Construction on Electro began in 1937, with his first major outing being the 1939 World's Fair in New York. Westinghouse had a whole pavilion built to show off their home of the future, and Center Stage was reserved for Electro. We have some videos from that first outing which gives us the full image of the machine. You see, Electro could be a little sassy. This was all part of his appearance as a non-threatening metal oaf. Here's some audio from that demo. You see, all I need to do is to speak

into this phone, and Electro does exactly what I tell him to do. Sometimes. But I don't see why I'm telling Electro's story when he's perfectly able to tell his own. So let's listen and see what Electro has to say to us today. All right, Electro, will you tell your story, please? Who, me? Yes, you. Okay, toots. Ladies and gentlemen, I'll be very glad to tell my story.

I am a smart fellow, as I have a very fine brain of 48 electrical relays. It works just like a telephone switchboard. If I get a wrong number, I can always blame the operator. Thank you. And by the way, I see a lot of good numbers out in our audience today. Electro, be yourself. Quiet, please. I'm doing the talking. I'm sorry.

You can actually hear in that demo the kind of affected speech that has to be used to talk to Electro. It's, I think, fascinating. But the bottom line here is that Electro was cool. He was the future, but he was an approachable future. He had a little bit of tude. This was only possible thanks to the earlier work done at Westinghouse. Once 1943 rolled around, Westinghouse moved into full wartime production. Luckily, Electro didn't get scrapped. Part of this is probably due to the fact that Electro is composed of aluminum and not steel. So Electro languished in a crate until one of the engineers who had built the machine took him home. Willy Vocalite became a weapon of war, but Electro was able to remain a pacifist.

A localite became a weapon of war, but Electro was able to remain a pacifist. Now, we're well out of the technical aspects of the story, but I do want to relate one final tale of Electro's wonder. He wasn't just the wave of the future, he was also a movie star. This kind of brings us full circle and back to the mechanical Turk level of illusion. After World War II, Electro was returned to Westinghouse. Over the next few years, he was restored, upgraded, and hit the demo circuit again. Then comes 1960, the year Electro went to Hollywood. He played a supporting role in a film called, and I'm not joking here, Sex Kittens Go to College. Now, dear listener, I've watched this masterpiece in preparation for this very moment. It's one of those classic

so-bad-it's-almost-good type of B-movies. It's also one of the worst films I've seen in years. type of B-movies. It's also one of the worst films I've seen in years. It's pretty explicit at points, so if you must watch it, then make sure there's no minors in the room. The plot is paper-thin. It has to do with a brilliant scientist who gets hired by a university. When she arrives, the faculty can't believe how attractive she really is. Hilarity ensues. Nothing too groundbreaking. Electro plays a character called Sam Thinko, and he makes up the strange bee plot. One of the college's faculty ran an experiment to see if Sam Thinko could predict things like horse races, lottery numbers, and the outcome of bingo matches. Turns out that Thinco is uniquely suited to the task. The professor leaves it at that, a neat experiment that once again proves that robots are truly the wave of the future.

But things go wrong. Someone has been using Sam Thinco's predictions to win bets. Thinco's so good at gambling that it's running bookies out of business. So a pair of mobsters show up to teach Thinco a lesson. Once again, hilarity ensues. I won't spoil the reveal. You'll have to see the film for yourself if you can bear it. However, there is a relevant scene near the end. Throughout the film, it's talked about that Electro has been blowing tubes, which is unusual. What no one knows is that Sam's been under a lot of stress. He's worried over the gambling scheme. Eventually, the mob catches up to him, and in a dramatic

confrontation, Thinco fries his transistors and starts smoking. He even has to be nursed back to health. There's this strange human quality to Electro's performance in Sex Kittens Go to College. He's not the villain or some tool of destruction. He's not even just a static machine. He's a robot that's in a little over his head. He even has a nervous break or something close to it. Sure, his brain is bigger than yours, but not in a malicious way. Thinco is depicted as lovable,

useful, and brilliant all at once. This is all thanks to the power of illusion. We should all know by now that Electro wasn't able to predict lottery numbers. He wasn't even computerized. He couldn't actually do math. He couldn't even speak on his own. But with a little set dressing and the right aluminum shell, you couldn't help but love the guy. Alright, this concludes Spook Month.

I'd like to end by proposing that robots can, in fact, be spooky. But all of those frights come down to context. Robots like the Mechanical Turk, which actually was a fraud, did really scare some people. Conversely, the Westinghouse series of machines had the opposite effect. Deliberately hokey design and favorable media coverage primed the public to fall in love with these weird robots. Scary or friendly, all these early robots were made possible thanks to the power of illusion. In the case of the Mechanical Turk, this was all classic misdirection. Von Kimplin built and operated the machine specifically to

trick people. The entire device was designed from the ground up in service of this overall illusion. It looked for all the world like a machine that could beat humans at chess, and even carry on a conversation. Some believed it was controlled by Satan himself. You could even look inside, under controlled circumstances. Audiences ate it up, how couldn't you? It was a miracle of modern science, or at least that's what it looked like. Westinghouse took kind of a backwards approach. Winsley was kind of forced into the whole robot thing, but he decided to embrace it. It's clear to see how

unsophisticated Herbert Televox was, but that didn't stop people from getting excited. Sure, it wasn't doing any great feats, at least not yet. In the very near future, this Televox thing was going to revolutionize the world. Herbert was just the first example of what a machine could do for us. Wensley never fully fell for the coverage. He always hedged by explaining just what his creation was capable of. The illusion was just giving people a little bit of what they wanted. The trend continued into Willy Vocalite and eventually the beautiful Electro. The technology itself didn't exactly change the world, but it did spark the imagination. Now, there's actually a lot

more to say about the robots of Westinghouse. I'd recommend checking out a copy of, well, Robots of Westinghouse by Scott Schout. It's a fantastic resource full of copies of sources and images of those strange robots. Let's close on a happy note. Electro is still very much with us. He's among the living, if you can believe it. Things got close for a while, but the marvelous mechanical man does still exist. He's currently on display in the Mansfield Memorial Museum in Mansfield, Ohio. Shout, the director and curator of the museum, was able to track down and reassemble parts of Electro. It's thanks to his work and the foresight of a handful of Westinghouse engineers

that we can still enjoy Electro to this very day. Thanks for listening to Advent of Computing. I'll be back in two weeks time with more of Computing's past. And hey, if you like the show, there are a few ways you can support it. If you know someone else who'd be interested in the podcast, then please take a minute to share it with them. You can also rate and review on Apple Podcasts. If you want to take your support up to another level, you can support the show directly by buying merch or supporting us on Patreon. Patrons get early access to episodes, polls for the direction of

the show, and bonus content. I'm revving up for a new bonus episode in November, so now's a good time to get in. You can find links to everything on my website, adventofcomputing.com. If you have any comments or suggestions for a future show, then go ahead and get in touch. I'm at Advent of Comp on Twitter, and I always love to talk to listeners. And as always, have a great rest of your day.