# Could keyboards + mice ever really become obsolete?

Ever since the keyboard/mouse combo really took off in the early 80s there has really been no satisfying replacement since--at least for complex/precision tasks such as computer programming and graphic/industrial design.

Often time computers in fiction are depicted with some kind of neural interface. While that seems like a cool idea, the mind is very impulsive, while you physical actions are obviously more deliberate. You often want to think something in your head before you actually do it--especially if your commands have real time consequences, like an air traffic controller for example (it's not all voice).

I do occasional Python scripting which involves automating various from desktop software . It's relatively simple but I cannot comprehend any system that could allow me to replicate this work efficiently other than at least a keyboard. Will we still be programming on keyboards 200 years from now? Even if it is some kind of laser/hologram/augmented reality keyboard? Using your finger as a "pointer" seems like a logical replacement for a mouse but in no way is it able to match the precision of a mouse or trackball, etc.

I'm probably just a product of my time, but I'm curious if other world builders have made any conclusions about this. And yes, this is probably slightly opinion based but not so much as that a logical hypothesis could not be reached.

• Do you ask about a keyboard as a concept (pressing some kind of buttons with your fingers), or as a dedicated mechanical device with fixed buttons, like modern keyboards? – enkryptor Oct 20 '16 at 12:03
• @enkryptor I suspect RTS doesn't mean Radio Television of Serbia – user27795 Oct 20 '16 at 14:00
• – user27795 Oct 20 '16 at 14:54
• Can you clarify whether "mouse" is to be interpreted as a separate physical mouse or whether your intent is to include laptop equivalents (nubs/trackpads) in that statement? FWIW, while other options can (and do) exist, I don't think there is going to be a more efficient and precise general input system. Task-specific inputs may be better at what they do (like game controllers, or touch-screens for applications optimized for touch-screens), but if flexibility and depth are needed I think you're stuck with the mouse (incl. trackpads/nubs, though less effectively) and keyboard. – GrinningX Oct 20 '16 at 17:03
• I would disagree on the "impulsive mind" problem with neural interfaces. If the interface is connected to the motor neurons, using the interface would be exactly like a motor skill - like typing - in accuracy and learning. – DOS4004 Dec 10 '16 at 4:28

I'm going to disagree with what seems to be the prevalent opinion: No, the keyboard will not become obsolete. The mouse, perhaps, but not the keyboard. It may change in its design (different material, different key placement, etc.), but as a big board of physical buttons, it is here to stay.

Why? Well, what would have to change for a keyboard to become obsolete?

There would have to be a simple way to enter complex data. This doesn't exist today; even if there were an interface today where I could simply think and the text writes itself on a screen, I would still program slower than I would with a keyboard, because thinking "for (int i = 0; i<=10; i++) {" is slower than typing it. "Right open parenthesis" is a lot longer to say or think than pressing shift and 9.

That means there will have to be new programming languages that either use natural language, or some kind of "virtual code" utilizing pictures or other representative objects in a virtual world. While there are some non-language-based forms of programming out there, none have come close to eclipsing written languages today. The software and hardware to support this "virtual code" will have to be developed to the point where "virtual code" is better, faster, and cheaper than its predecessor, because otherwise, what's the point of switching?

Even then, there will be programmers who will happily enter this virtual world and program themselves a "physical" keyboard. Remember, people still program in COBOL - just because a replacement exists, doesn't mean people will stop using it. Even a purely virtual world has a use for a keyboard.

All that is a long, long way away, and as long as keyboards are cheap, intuitive, responsive, and touchable, I doubt that time will ever come. Keyboards may evolve, but they will never go away.

• I really like the concise argument for "virtual code". As well as hitting on the economics of keyboard being cheap. That is definitely key. – Cradle2theGabe Oct 26 '16 at 16:54
• Good answer except "Never" is too long. They'll go away if/when we all have direct computer-brain interfaces. IA rather than AI. It'll happen sooner or later provided our civilisation does not crash and burn first. Doubt it will happen this century, though. – nigel222 Dec 10 '16 at 12:33

They're already going away.

There are currently around 2 billion personal computers in the world (the kinds of things with keyboards and mice). In contrast, there are about 6 billion devices connected to the internet not including personal computers. Devices that don't have a keyboard and mouse outnumber those with by 3-to-1, and we haven't yet included computers that don't connect to the internet.

Current examples of alternative methods of input for computers:

• Video game controllers (which can even have well-designed alphanumeric input; see Steam Big Picture)
• Your car's steering wheel and pedals
• Everything you do on your smartphone
• And tablet
• Lots of custom inputs (e.g., coffee maker, refrigerator, Nest thermostat)
• Voice (Siri, Cortana, Ok Google, Dragon)
• Three dimensional holographic manipulation (consumer ready now? No. But I've seen the concept designs and prototypes, and it's pretty great)

Okay, okay, you specifically asked about "complex/precision" tasks such as programming and graphic/industrial design.

Graphic/industrial design is easy: you want a pen-enabled touch screen. A big one. Pens not only give you pixel-perfect precision, but removes the barrier of the mouse concept for a more direct interface.

As a software engineer, I really can't imagine anything better than a keyboard and mouse for programming. That being said... I've been using them for so long, that at this point it's just a personal block I have. As an author, I can't imagine anything better than a keyboard and mouse for writing, yet Terry Pratchett wrote most of his final books with his voice. While it would certainly require a very special programming language and IDE, I could certainly see people using voice to program (other people, not me; I really can't get into voice controls for my GPS, let alone programming).

But, about that mouse... I haven't seen a desktop computer in quite a while. And if we're talking laptops, pretty much all laptops come with either a touchpad or the "nubbin", and not a mouse, to control the on-screen pointer. Many people no longer bother with a proper mouse anymore. And if you're concerned about speed and precision with either of those, that just means you're not an expert in them; people who do use them exclusively are just as good as you are with a mouse.

Some people don't even use mice. I've even heard arguments that the mouse/keyboard combination is slower than pure keyboard input (though I'm sure that's entirely dependent on the person).

The most interesting barrier seems like it's going to be how to get large blocks of text accurately into a system. An argument could be made about the difficulty of speaking special characters for programming, but I don't believe that to be true. If we had, for instance, a pure voice method of programming, we would design programming languages conducive to that format. After all, if Ook! is a language...

Anyway.

Could voice replace hands for lengthy text inputs? It certainly seems possible. Speaking is something that basically everyone does, while keyboard use is driven entirely by technological restriction and necessity. As voice controls get better, fewer and fewer people will use keyboards (other than specialists). And once voice gets good enough, even the specialists will probably switch over.

• "As a software engineer, I really can't imagine anything better than a keyboard and mouse for programming." I can: a keyboard. With an emphasis on the absence of mouse. That said, a SWE interface to a phone is very intriguing. – user58697 Oct 20 '16 at 3:55
• I don't know if "going away" is the right term here. Yes, they are in the minority for devices connected to the internet, but that doesn't mean that the other options are superior. I spent almost two hours today programming via a touch screen interface. And it's awful. No shift, no control, no arrow keys, no tab, no quick access to multiple symbol characters, such as { and } (which you kind of use a lot). NO SCROLL WHEEL (and no, I did not have "swipe to scroll"). Then you have the fact that M+K is just hands down better at first person shooter type games than controllers. – Draco18s Oct 20 '16 at 5:53
• I can't agree a touchscreen is an alternative for keyboard. When it comes to typing text, all the touchscreen does is basically mimicking a keyboard. The concept itself is still the same. – enkryptor Oct 20 '16 at 11:47
• @Azuaron Ask them to move something over by precisely 1 pixel with their pens. :) – SPavel Oct 20 '16 at 15:37
• «Ask them to move something over by precisely 1 pixel with their pens» just as bad with a mouse! I would use a keyboard nudge feature, or use the x/y properies and type the number. I use pen and trackball and jog/shuttle and special buttons. – JDługosz Oct 20 '16 at 22:00

## Of course they will become outdated

### Limitations of Mouse and Keyboard as Input Systems

Think about what the primary purpose of the keyboard and mouse are:

It is to serve as a low-bandwidth intent-to-action converter (keyboards can accept about 500 keystrokes per second, but our own physical limitations mean the average input is far below that). The low-bandwidth can be partially mitigated by strategically designed shortcuts and symbolic logic -- witness an experienced programmer blasting through console commands, or a travel agency booking agent using those 'outdated' green screen at speeds GUI users could hardly dream of. The disadvantage is that any user who hopes to use the system must learn the symbolic logic, which can take a lot of effort. All that info has to be 'cached' in the user's mind.

A mouse (and its touchscreen equivalent, the tap) goes the other way. The mental 'cache' goes away, and low-level widely-shared symbols, such as windows, icons, buttons and selectors take their place. This is still is another intent-to-action converter. The screen is a (one-way) high-ish bandwidth connection taking advantage of the fact that 20% of our brain processing power consumption is dedicated to visual signal processing. Megabyte one-way communication is useful, but a mouse is basically a binary selector -- I am interested in this area of the screen -- tap/click. All interaction has to be structured as a series of branching screens. Moreover, the megabyte connection overwhelms our (relatively limited) 8-bit or so attention channels - witness the difficulties in making websites and software useable, whereby useful information is often missed by users.

### A Path Forward

If you think about it, human interaction with computers has moved through various stages of abstraction, as we have designed ways of interacting that are more and more human-cognition friendly. Initial communication with machines was at the 0/1 bit level, and this was so difficult that only math Ph.D.s could write compilable programs. This was replaced by machine language, printers, displays, programming languages and finally user interfaces. A ten-year old can now initialize a neural network by clicking a few buttons.

How can machines get even better at reading human intent? I can think of two ways:

## The Butler Model

### Embedding machines into the fabric of our world. Effectively this provides the machines with a tremendous amount of information on our lives and wishes, in other words context, a "common sense" of sorts and makes them more likely to guess what we might want, given our prevalent conditions.

Heart rate is irregular, such and such enzymes are building up, present user with this specially designed enzymatic drink to bring them back to nominal parameters.

It is 6pm, user is closing her drawing pad at work and wrapping her things, probably leaving for the home arcology --> call a vehicle.

## Subconsciousness Model

### Embedding machines in ourselves -- mind-machine interfaces that read our intentions directly can then convert those "intents" into actual machine code, without any conscious direction by us.

Alice sits in the middle of her virtual playground. The world shifts to become green and hilly. Streams erupt around the valleys, the world quickly adjusting to look as if the streams had eroded the valleys over centuries, plants shift to meet Alice's implied climatic preferences. A castle emerges in the distance, first with Disneyesque towers and crenels... Alice starts to frown ... and the castle shifts to a darker, gothic style. The sky darkens a little. Forests pop out in the mid-distance, and snow covered mountains become visible in the distance.

In orbit around Venus, Ahara decides to build a 100km habitation module, and spend a few days in the Venusian clouds. Her holovision creates a mockup, with the best guess of the machines often adjusted slightly by Ahara's thoughts. The machines take care of hull integrity, propulsion systems, optimizing across over 200 dimensions. As she thinks, the 3D printers create, dismantle and recreate the module live and so quickly that Ahara never bothers to think which parts are holovirtual and which real. Within minutes, Ahara climbs in and jumps into the Venusian inferno.

I think the problem with many suggested improvements, like touch screens or mind-machine interface, is that they're more effortful than the status quo. And voice recognition? People talk too much anyway. The spoken word has a low degree of precision too. A high precision mouse allows you to move your hand very little to produce great change inside the computer... rather than throwing my hands all over a big screen to do the same thing.

With a mind-reading machine you'd have to focus quite a lot on what it is you are doing, but with a little experience a mouse and keyboard become instinctive movement to translate your intent into action. And, importantly, you can be clicking away at whatever task while thinking about something else. That isn't something which can as easily be done with a mind-machine interface. What if you get distracted or are thinking about the next step?

What I suggest is that we take the concept and improve on it in terms of precision, efficiency, and customisation. Replace the mouse and keyboard with holographic equivalents, which can be modified considerably for increased efficiency. Like, automatically scaling them for a person's hand and finger size and shape, and moving them closer to their body for ergonomic ease.

Then you have the opportunity to interact with, let's call it a keysphere. You can design the holographic interface to be take advantage of 3D space and ergonomics far better than a flat board. The shape of the keyboard, and the exact gestures the mouse uses, can change to suit whatever objective. The mouse in this case becomes more flamboyant too, and may become more like a holding an imaginary ball and twitching it to do more complicated tasks at once. Plus more hygienic too - you're not touching anything! And let's not forget the pretty glowy holograms! Oooh. Holograms.

If by keyboard you mean the standard QWERTY keyboard as found on contemporary machines, then sure, they'll change or fall out of favor.

But if you mean using fingers to press keys or buttons as a way to input information, no. That's not going away.

As long as human beings rely heavily on their hands and fingers to interact with their physical environment, some sort finger-pressing operation will be involved in almost all interactions with mechanisms.

We don't use keyboards because nothing better has been invented yet, we use them because we have thousands of years of using our fingers to do almost everything that matters for our survival, and our bodies don't have any better, more efficient way to translate our thoughts into physical actions.

Our brains are incredibly good at making our hands and fingers do complex tasks very quickly and with minimal physical and mental effort.

Well, as we change so do our tools. Going with this i believe that ,while we stick with our current media, we may be stuck with the mouse and keyboards. However,still we can change the designs of the keyboard. For example, take a flip phone, it has letters divvied up between 9-12 keys,and you press them a certain number of times depending on what you want to type. going with this, we could create something similar. If you created a 12 button keyboard, with only only 3 having functions other then letters, you could create something more efficent;it might look like this (excuse my terrible art-talents) The red keys would represent normal typing ,the yellow would sigal being done with a letter selection, the pink would be space, and the green would be tab/enter. thats all i got.

Im going to add something to the answer of Azuaron:

You can see today a decrease in the market value of desktop pc's and laptops. A increase of other technologies will follow with the newest being VR Headsets that can be applied to practically anything, Now imagine a good vocal recognition software! You could do anything with those little headsets and the market is booming or is going to boom soon, the stats say so.

If you look at the statistics of the market size of VR headsets you see a exponential increase until 2020. The technology isnt only interesting its also supported financially through many people.

I think the mouse is going to become obsolete soon but the keyboard is, as Azuaron says, still very important. I program myself and a colleague here uses his mini 13'' without a mouse. The funy thing is, he is faster than most of the people I know. Eventually voice recognition will become just as useful as the keyboard.

• Have YOU ever tried to talk for a whole 8-hour work day? Yes, there are a few people who can do it, but such compulsive talking is sometimes considered a mental illness, See logorrhoea. – jamesqf Oct 21 '16 at 5:59
• Let ME be happy about the newest technology. Considering self driving cars and huge advances in medical technology in the last years (see: ethz.ch/en/news-and-events/eth-news/news/2016/09/…) we can take into account that there could be systems that would be controlled only with our imagination. – Fenrir Oct 21 '16 at 7:22

Quick answer - yes they'll go.

Slow answer - can you imagine them still being a major means of digital interaction with whatever computers have reached in say, 300 years? I can't.

A lot of the above answers seem to assume today's outlook - we are used to keyboards and they are useful for precision tasks.

I think that, given another generation or so, especially one that takes better neural interaction in some form, for granted, and the effort and idea of typing letter by letter will be something quaint that only granddad's (and grandmums) do.

You think we will never develop direct neural interfaces???

For kingledion: You really think mice and keyboards would survive in a world where we have a direct mental connection to the computer? Why use the brain to command our hands to command a mechanical device to command the computer when the direct connection is better?

The monkey equipped to directly command a robotic arm learned there was no need to move it's physical arm to command the robotic arm and quit doing so. Why would humans do it the hard way when the easy way is something a monkey can figure out on it's own? (The researchers set it up so the robotic arm would behave as the monkey's real arm, but the connection was in the brain, not in the arm. Even though they weren't intended to be separated the monkey figured out how to simply think at the robotic arm without moving the real one.)