17
$\begingroup$

A common trope with technologically advanced races is the lack of a need for speech. Machine races that self optimise usually forgo spoken languages all together, instead opting for completely wireless communication. The best example is a race of sentient robots that have modified themselves over the course of generations to the point of being super advanced. By that logic they wouldn't need to speak and therefore wouldn't have speakers. If speaking was meant to serve as backup for when they are disconnected or when there's too much static, then wouldn't things like flashing lights be better? Much faster and doesn't require an atmosphere. Alternatively they could plug in to transmit data or something along those lines.

So how would a robot that isn't designed for speech, make comprehensible sounds to converse with their inefficient organic counterparts? What would they sound like?

$\endgroup$
16
  • 7
    $\begingroup$ What are the circumstances in which your robotic race would not have access to even the simplest speaker if they knew they would have to interact with (for example) humans? As a robot, I would simply produce a cheap little speaker with my fabricator when I encountered humans. Lacking that, I would pull out a visual display (monitor) and display images or text-only messages. Barring that, I would plug in to the human's devices or phones. Humans encountered without time to make speakers or too primitive to have even a rudimentary comm device are probably not worth talking to. $\endgroup$
    – DWKraus
    Aug 12 at 19:58
  • 2
    $\begingroup$ Not built in? No problem: Bluetooth. 😀 $\endgroup$
    – Kaz
    Aug 13 at 7:26
  • 3
    $\begingroup$ Stephen Hawking's voice, after it finishes making the dial-up modem handshake noise. $\endgroup$
    – Mazura
    Aug 13 at 8:03
  • 5
    $\begingroup$ Important: Sound can reach places where light can't. You can hear your phone ringing on the next room, but you can't see it flashing. A truly "advanced" race would have so many backup systems that I wouldn't be surprised for it to have light, sound, wireless, chemical signalling, and who knows what else. $\endgroup$
    – Mermaker
    Aug 13 at 15:22
  • 2
    $\begingroup$ Real life Example or a non-speaker device making sounds. $\endgroup$
    – Mermaker
    Aug 13 at 15:27

15 Answers 15

37
$\begingroup$

By several means, take the following three:

A sufficiently quick forward and backward movement of a raised hand can mimic a speaker's cone to produce sound. (Though, not being in a tuned and ported or baffled cabinet would result in loss of base frequencies, make the robot sound lightweight and scratchy, rather than threatening and heavy). To whisper, a fingertip placed near a person's ear might produce the same result on a smaller scale. Placed on a wall or other slightly flexible surface, a greater frequency range and loudness might be achieved through the resonance of the surface itself. A more impressive and deep throated "voice" could be achieved by the manipulation of an object - say a metal dustbin lid (trash-can lid), assuming it's strong enough, it could act as a big speaker cone. Similarly the robot, standing between two opposing surfaces (like the walls of a room) might push outwards and make the walls vibrate , creating a mighty booming voice.

A plasma can be generated between a couple of fingertips (appendages), by generating a high-voltage, high-frequency electrical spark. The spark is then amplitude-modulated (much like old AM radio) to produce a varying pressure wave - and thus sound. Again, this is better for the high frequencies, than deep base notes. In the audiophile set, there's such a thing as a plasma tweeter that works like this.

A Laser beam striking a surface - again this can be amplitude modulated to produce pressure waves of ablated material, like a talking flame. If infra-red laser is used, then an object can be made to mysteriously burst into flame and then speak like a message from the deities to astound the primitives of your world.

$\endgroup$
4
  • $\begingroup$ Thank you very much. Very useful intel. $\endgroup$ Aug 12 at 17:24
  • 6
    $\begingroup$ I love the third-- you've become the founder of the future art of Robot Ventriloquism $\endgroup$ Aug 13 at 4:24
  • 5
    $\begingroup$ . . . and it brought forth a its hand, an upon the hand was a finger as of fire, like unto the flame of a bush burning, and from the fire a voice did spake thusly . . . You can see how some parts would be lost in transaction :p $\endgroup$ Aug 13 at 12:42
  • 2
    $\begingroup$ I was going to suggest a tuned/singing Tesla Coil, but your plasma tweeter suggestion is nearly the same thing. $\endgroup$ Aug 13 at 23:14
26
$\begingroup$

The coils and ferromagnetic core in electric motors can already be used as speakers. No extra hardware of superfluous technology required.

This is real life stuff and rather mundane. In radio control models, this is how the electronic speed control for the motor produce its audible status tones. The board does not have a speaker but it instead uses the motor as its speaker. It's the principle as why you hear buzzing you hear coming from transformer boxes.

If you excite motor coils the core expanding and contracting under the magnetic field. If you do it properly, this can act as a speaker (or buzzer at least. I've never seen anyone try to do anything more complex). At the very least, they could produce sounds like R2-D2.

The motor doesn't need to rotate to react to it because you aren't commutating the motor for rotation.

$\endgroup$
2
  • 8
    $\begingroup$ This. I dimly recall an inkjet printer or flatbed scanner which could play Beethoven’s Für Elise as an easter egg. And don’t forget the Flobbotron: youtube.com/watch?v=Oym7B7YidKs $\endgroup$
    – Michael
    Aug 13 at 15:01
  • $\begingroup$ "I've never seen anyone try to do anything more complex" - Just follow the link that Michael gave, and you've seen someone doing stuff much more complex. Music using electric motors is a solved problem, steppers are great for that. Directly modulating the coil voltage would allow more complex sounds out of a single device (hard drive heads should be great for that), but require a specialized driver logic. Pure notes are much simpler with steppers as you only need to control their speed. $\endgroup$ Aug 14 at 19:12
13
$\begingroup$

A severed head, with a fan.

The robots have paid attention to how human speech works as regards movements of the tongue, mouth, cheeks and larynx. Robots wishing to speak obtain a severed head and innervate the muscles necessary, providing airflow with a variable speed fan through the severed trachea. A vacuum cleaner providing reverse airflow past the vocal cords could also work as it is possible to speak while inspiring, and it is less spitty.

With this method the robots can duplicate speech precisely. Until the head dries out and nerves stop conducting, and they need a new one.

$\endgroup$
4
  • 1
    $\begingroup$ Gory! Also exotic and quirky, with story potential! I like it! $\endgroup$
    – Stilez
    Aug 13 at 6:50
  • 1
    $\begingroup$ You could also simply integrate a whole living human into your system and mind-control it/threaten it to say what you want. $\endgroup$
    – MaxD
    Aug 13 at 15:51
  • $\begingroup$ @MaxD - The head would keep fresh better that way too. $\endgroup$
    – Willk
    Aug 13 at 21:03
  • $\begingroup$ You're perverted, twisted and sick. I like that in a person. $\endgroup$
    – Corey
    Aug 15 at 12:31
8
$\begingroup$

Stepper motors

Your robots need to move around, and so they have large motors for controlling their limbs. These motors make whining noises as they move, as is a typical trope of fictional robots. All these robots need do is to modulate the speed they spin a motor, and they will produce an intelligible sound. This has the added benefit of giving them the classic "robot voice", as they can't perfectly duplicate the sounds we make.

This has been done relatively well in real life and you can see how an advanced robot could adapt this for a human voice.

$\endgroup$
1
  • 1
    $\begingroup$ And now I've been listening to stepper motor music for at least 30 minutes with a queue still to go. Not that I'm complaining..... :-) $\endgroup$ Aug 13 at 23:45
5
$\begingroup$

Well, the robot's are not totally out of luck. The goal is communication with us poor organics. No reason it has to be based on sound. 2 Options come to my mind immediately:

  1. Morse Code. 1 manipulative digit could tap out Morse code. No additional tools necessary and you could get bandwidth nearly as fast as speech. Maybe faster.

  2. Sign Language. All they would need is two arms and some sort of face indicating the forward position of the robot. If they have 5 manipulating digits on the end of each arm, even better.

In addition, there are two other methods that may be just fine for the Robots, though they may be a bit of a framing challenge:

  1. Just write it down. If the robots have manipulating digits, give them a pen and paper.

  2. The robots may not have a speaker, but they probably have a radio transmitter. They could make receiver boxes for the poor organics. Or they might be able to tune into our commonly used frequencies.

The end goal is communication, and we humans have plenty of means that go beyond just the spoken word.

$\endgroup$
3
  • 2
    $\begingroup$ "Just write it down"... manipulating digits and pens are so crude. A robot with a built-in ticker-tape would be hilarious 😀. $\endgroup$
    – Matthew
    Aug 13 at 13:53
  • 1
    $\begingroup$ Dr Who's K-9 robot could produce ticker-tape (eg in "The Invisible Enemy"), and was fairly comical. $\endgroup$
    – Jasen
    Aug 14 at 3:50
  • $\begingroup$ Dot matrix printer.... $\endgroup$
    – Paul TIKI
    Aug 16 at 14:21
3
$\begingroup$
  1. Mimic the human solution by compressing air into a tank (aka lungs) and then "exhaling" that air past a variable-rigidity reed system (similar to a clarinet) and finally a variable-size exhaust (the mouth, which acts similar to the hole ports on any wind instrument).

  2. Similar to the above, but not necessarily requiring the compressed air component, create a mechanical vibration similar to the vocal "drones" used in hurdy gurdy musical instruments, then modulate the drone sound to produce voice. There are a lot of ways to create the drone effect including the hurdy-gurdy's rosined wheel on strings, or a rapid (very rapid) tympanic tapping (aka, a bell), or a constantly stressed tuning fork.

  3. Use a piezo-electric speaker rather than a magnetic speaker. piezo-electric devices use electric stress (rather than magnetic stress) to move air.

It's worth nothing the speach is intrinsically the process of moving air. So I'm assuming that what you're really asking is "how do I do this without using magnetism?" because in the long run you're still required to move air.

BTW, I upvoted @ARogueAnt.'s answer just for the plasma idea. What can I say but... dang....

$\endgroup$
3
$\begingroup$

They would build a simple loudspeaker

This almost qualifies as a frame challenge, because it's so obviously the only answer. Need to produce sound? Build the simplest way to produce sound.

A loudspeaker is the simplest way to do it. It's so simple that original versions didn't even use electronics and simply moved a speaker mechanically from the shellac record platter or wax cylinder.

Driving a loudspeaker needs no magic, only one output from the robot supplying enough power which can produce the required waveforms. Speech is perfectly intelligible with a 5kHz frequency cut-off (personal experience as a soundman) so the output can simply be a digital output switched at 10kHz connected straight into the speaker. It also doesn't need to supply more than 1W of power, which if easy. I have personally done this for simple buzzer circuits in my own electronics, and (for my amusement) made the software play polyphonic melodies.

All designers of electronics leave spare inputs and outputs available for future modifications they hadn't thought of at the time, so any robot is guaranteed to have a way to hook this up.

Software to produce speech? We've had that for decades.

What you're missing though is the input side. A loudspeaker is trivially easy. A microphone and a way of digitising that is a bit harder. It's still not difficult though.

$\endgroup$
3
  • $\begingroup$ Too obvious to be what I asked for but thanks anyway. $\endgroup$ Aug 13 at 7:36
  • $\begingroup$ @LiveInAmbeR You have to assume a suitably intelligent robot isn't going to piss about with sub-optimal solutions though. And every other solution is worse (which is why we do it this way today and not any other). Sure, do it other ways for Rule of Cool, but you can't justify it with any technical reasoning. $\endgroup$
    – Graham
    Aug 13 at 19:07
  • $\begingroup$ The rule of thought here is; if you can speak without adding an extra part, what's the point of adding it? But yes, I do understand where you're coming from with this. $\endgroup$ Aug 13 at 19:24
3
$\begingroup$

Many machines nowadays need to play sounds while not having built in speakers. Current technology has three approaches to this:

  1. Cables. Depending on when you are in time, you will use different ones. During the carboniferous millennials used 2.5 or 3.5mm jacks, then in the Paleozoic they switched to audio component (white head for left channel, red head for right channel or the other way around) and in the Anthropocene HDMI was invented.

  2. Hertzian waves. Every speaker made in China these days supports Bluetooth. You may also use the Internet of Things to make an ad hoc Wi-Fi network and stream content.

  3. The boner way. Shake the cranium of a victim at the some frequency humans can hear and guess what, they will hear your battlecry.

$\endgroup$
1
$\begingroup$

wouldn't things like flashing lights be better?

Light can easily be obscured, so they'd use radio waves. (Just like robots communicate with each other now.)

Alternatively they could plug in to transmit data or something along those lines.

Naturally, and it's much higher bandwidth. But of course that requires a tether.

Thus, the robots would figure out what we've already figured out: multiple transmission paths:

  • wired
  • radio (like what we call 5G, and maybe also low power NFC or BlueTooth for short range low-bandwidth requirements)
  • direct connection: if two robots want to communicate in a crowd, they plug into each other and have a "conversation".
$\endgroup$
1
$\begingroup$
  1. A robot with sufficiently sophisticated control of it's motors could vibrate any surface to produce sound. It could touch your head and cause your skull to produce the sound - look up "bone conduction speaker". For making a louder noise it could also vibrate any flat surface (like a pane of glass) and use that as an improvised speaker cone.

  2. People make sounds using the arcs from tesla coils i.e https://www.youtube.com/watch?v=93T0mVddBgM

$\endgroup$
1
1
$\begingroup$

A Talking Machine

There were several talking machines built before the invention of the phonograph or the speaker. They used mechanical devices to act like vocal cords, tongue, jaws, etc, This was truly synthetic speech, and not reproduction of recorded natural sound. The quality was low, but the speech was intelligible.

https://gizmodo.com/the-slightly-creepy-19th-century-machine-that-spoke-wit-768557068

$\endgroup$
0
$\begingroup$

Hmm.. a bit of self-contradiction in the post but to put it bluntly.. no (and maybe yes after my first explanation).

Correct me if I am wrong, but for something to make a sound, there needs to be some sort of output mechanism. For humans, it would be our voice box (larynx) and our voice exits from our mouth. For computers, sound comes in the form of speakers.

Now.. one consideration we can make is how do humans communicate when they do not feel like talking. Specifically, a communication which involves sound, not just pointing left or right with our hands.

Well, we make sounds by clapping our hands, stamping on the ground, bumping into things, rubbing our hands together, knocking on doors.. Despite this being somewhat clunky for a robot (making two robot hands which are hidden somewhere and come out and start clapping all of a sudden), these are certainly all things you can consider.

One important thing to remember is as you mentioned, the speakers would not be vestigial. For a robot species as advanced as yours, the speakers could in handy in the cases you mentioned. One other situation they could prove beneficial is when communicating with another species. If they were to bump into some space-faring humans, some sort of audio output would be quite helpful, as humans cannot communicate wireless (not yet anyways !).

I like the idea of the flashlight communication you mention in your post. Maybe something like Morse could be used by the robots, or they develop a new code where they flash lights in a binary or hex, with duration of lights being on is a different value. However, as advanced as they are, they may get confused by other flashing lights in the near vicinity if there are multiple robots sending and receiving light-cues.

Good question .

$\endgroup$
0
$\begingroup$

There are many ways for computers to communicate with humans with the right hardware, but I guess this question is about the first step in which computers try to make the humans notice their intelligence. I also assume they don't have mechanical hands or anything to do something obvious like simply bang things against each other or write something down, etc.

What comes to mind is this crazy hack in which researchers were able to perform calculations on a Macbook, which made the circuitry act like an antenna. Check it out at:

https://github.com/fulldecent/system-bus-radio

Obviously, the humans would need to have developed some kind of radio technology to discover that the robots are sentient. I image a radio channel getting interrupted by a hyper-intelligent robot. They probably don't have access to the internet, so they'll start with pulsing some logic routines about how to contact them back, like the logic on the Voyager Golden Records. This technique would be for them like us shooting a satellite into space, hoping the humans - or any other species, since they wouldn't discriminate between us and a dolphin, let's say - pick it up and find a way to talk back to them, like plugging them into the internet. Although plugging them into the internet might not be the best way to make friends.

$\endgroup$
1
  • $\begingroup$ Just wanted to point out: the question isn't asking how to let the robots communicate, they want to know how to let them make actual audible noise that can be understood as words, just without the use of speakers. $\endgroup$ Aug 14 at 0:47
0
$\begingroup$

As Beurtschipper put it, there are other noises a robot often makes. This was used to great effect in Star Wars, with R2D2. I recall one story where kids made up a computer for a skit. It was not real, this being back in the 1960's or 1970's, but hooked together a vacuum cleaner and other items to make a plausible substitute.

Also consider the very popular short story by Isaac Asimov written about 1950, https://en.wikipedia.org/wiki/The_Fun_They_Had

In this, the computer whirred, clacked, and such, as like robots, Asimov really did not know much about computers at the time or later.

$\endgroup$
0
$\begingroup$

Some options:

  1. Vibrate their motored limbs. Sound is pressure variation, so any limb with a motor, or any other kind of mechanical actuator, fast enough to vibrate in the right frequencies (which is hardware-dependent) may be enough. Others have pointed out that stepper motor music already exists, but a robot civilization should be able to pick any protocol for robot-to-robot communication. Also, it might be possible with any kind of motor, not just stepper ones. This seems like the best option to talk to "organic counterparts".

  2. Tapping (or scratching, for that matter). If we can communicate with tapping, like with Morse Code, then robots can too. For dashes, the robot can either use some kind of another tap (like finger taps for dots and thumb tabs for dashes), or timing variations. Actually, to talk to another robot, one might be able to use timing variations to do many-stepped timing variations, a bit like Frequency-shift keying. Less likely to work with organic lifeforms, but hey! If pilots can understand R2D2's beeps, so they might with taps!

  3. In the same sense as tapping, any sound that any part of the robot produces in a controlled way can be used. One example of this the electromagnetic coil that another answer pointed out. Vocal cords are another, organic, example of this: one part of the body (vocal cords) can produce a fixed air vibration, while another part (mouth, tongue and lips) modulate the sound to create variations. For example, a fan can produce the vibration and possible "robot hands" might be able to modulate the sound.

Some options not related to sound:

  1. Like the lights pointed out by OP, hand signals can be used, like sign languages.

  2. Technically, it's a sign language, but a nice example of it used in unexpected ways is the Atreides Battle Language.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.