The Keyboard is Dead Thread
This seems to be bubbling up in a few different threads, so I wanted to give it a home.
Sure, maybe it's a little early to say this although not much earlier than the "Floppy is dead" pronouncement that accompanied the original iMac.
Here's why-
1. Multitouch means Multiple touch responsibe input! NOT just multiple single touches.
2. Keyboards limit form factors - Apple hates that.
3. Typing is an inefficient means of data entry.
4. The iPhone will teach everyone a new way.
Here's what will happen-
1. The iPhone, either intrinsically or via a widget, will allow for text entry via chording.
2. Teenagers, already willing to 9key text messages, will readily learn the faster chording method.
3. Soon they will hook their iPhones up to computers to use as keyboards/trackpads.
4. Apple will release a tablet or - to dream - a double screened tablet (somebody do a mock up, please!)
If you're still not seeing it-
1. Translucent input overlays become like training wheels until chording is mastered.
2. Ambidexterous chording method allows for left/right, one/two handed entry.
3. No need for tactile feedback as you don't have to hit a specific spot on the screen.
4. Hot keys replaced with gestures.
5. No need to look at the screen.
Sure, maybe it's a little early to say this although not much earlier than the "Floppy is dead" pronouncement that accompanied the original iMac.
Here's why-
1. Multitouch means Multiple touch responsibe input! NOT just multiple single touches.
2. Keyboards limit form factors - Apple hates that.
3. Typing is an inefficient means of data entry.
4. The iPhone will teach everyone a new way.
Here's what will happen-
1. The iPhone, either intrinsically or via a widget, will allow for text entry via chording.
2. Teenagers, already willing to 9key text messages, will readily learn the faster chording method.
3. Soon they will hook their iPhones up to computers to use as keyboards/trackpads.
4. Apple will release a tablet or - to dream - a double screened tablet (somebody do a mock up, please!)
If you're still not seeing it-
1. Translucent input overlays become like training wheels until chording is mastered.
2. Ambidexterous chording method allows for left/right, one/two handed entry.
3. No need for tactile feedback as you don't have to hit a specific spot on the screen.
4. Hot keys replaced with gestures.
5. No need to look at the screen.
Comments
"The most significant disadvantage of the chording keyboard is that it cannot be used by an untrained person. At least 15 hours of training and practice are necessary to learn the chord patterns that represent individual letters and numbers. A second disadvantage of the chording keyboard is that data entry rates (characters per unit of time) are actually slower than data entry rates for conventional keyboards. Due to the increased learning time and slower performance, chording keyboards have not become commercially viable except for specialized applications."
In fact, a little Googling shows you've been on this chording crusade since at least 2004 here on the forums. Give it up. It's not going to happen.
This is wishful thinking based on questionable assumptions. What makes you think chording is faster than typing?
Ok, I also did a quick google and found the average typing rate is around 30 wpm. Going off the wiki on chording I found one research study that suggested one handed twiddlers get to 47 wpm.
No, I didn't bother to look at sample sizes or do a multi-study evaluation, but I think it's obvious that chording will be faster than 9key text entry.
As for whether two handed chording can be faster than two handed typing, I'll make a four part argument-
1. Why it must be at least as fast.
If you think about the typing you are already doing it isn't far off from chording. You may only extend one finger for a key, but the other fingers are assuming some relative position to the extended one as you do this. At the very least if you were typing on a screen, if the screen just kept track of the relative positions of your fingers it would mimic Qwerty in speed and functionality.
2. Why it can be more accurate.
Now imagine that instead of your typing being dependent on physically locating an exact key the keyboard paid attention to the relative positions of your fingers. Instead of determing the key you want to hit by what key is depressed it has five seperate pieces of data to judge (plus the relative position of the virtual key if you are using a virtual keyboard of some type), thus increasing accuracy.
In addition since you are essentially "typing by wire" the more judgements can be made by the computer as to what you are trying to enter. It can watch your words and if there is some ambiguity between whether you were going for an "e" or an "r" but the preceeding text was "Appl" the computer can lean toward awarding an "e."
3. Why it will be faster AND better.
Increasing the accuracy alone should increase speed, but you can also do some things with a touchscreen interface that cannot be done with a keyboard or twiddler. Consider gestures. Apple demonstrated that it can be faster to use two fingers to scroll than to grab a scroll bar or find a scroll button, the same can be true for text entry.
Also, a more accurate measure of an effective computer entry method wouldn't be based simply on words per minute, since we do a lot more than enter text. With your hands already on the screen you will be much faster at clicking, dragging, dropping, and generally interacting with our computer. No need to reposition your fingers on the keyboard afterward, you just start chording again.
4. Why it won't be that wierd.
We already chord - hot keys, shift, option, command, etc...
Finally, I'm not sure my message received such a confrontational response. This is an Apple forum. Ya know, think different and all. Maybe I was terse, but only because I wanted to succinctly address the major issues that seem to be getting in the way of people realizing the potential that is here. So what if I've been talking about this since 2004, there's people on this forum that have been hammering at the need for a new newton since, well, the newton and, guess what, they got it with the iPhone.
The only thing that scares me about the chording, touchscreen entry future is that it won't be thought out well and we will be stuck with a crappy system just because some company has market predominance *cough* Microsoft *cough*. If I was working at Apple I'd be hiring experts on kinesiology, carpal tunnels researchers, linguistic experts, and do a lot of trials and studies until I was sure I had the most easy to learn, efficient, accurate, comfortable system possible.
Apple has the opportunity to define the future of how we interact with a computer. I expect them to do it right.
This would offer a few advantages, though not exactly be a revolution. I think this video show a good example of such a technology (keep in mind this was BEFORE the iPhone):
http://www.gametrailers.com/umwatcher.php?id=23071
So you're not saying the keyboard as an input device is dead, just that it will become digital rather than analog. It will still be keys, and not something else (eg, voice input) but will be on a digital screen rather than physical keys.
This would offer a few advantages, though not exactly be a revolution. I think this video show a good example of such a technology (keep in mind this was BEFORE the iPhone):
http://www.gametrailers.com/umwatcher.php?id=23071
Does anyone on this forum ever use the search function?
There's already been a lot of discussion on the topic in several Appleinsider threads including these:
http://forums.appleinsider.com/showthread.php?t=71246.
http://forums.appleinsider.com/showthread.php?t=71252
http://forums.appleinsider.com/showthread.php?t=70695
There was a lot of discussion and speculation on Multi-Touch, chording, gestures, virtual keyboards and the now famous Jeff Han video cited in the previuos post by turnwrite. Kolchak is the resident forum expert since he's a long-time user of Multi-Touch via the old FingerWorks iGestures pad. You can see his posts in the previously mentioned threads.
Come on folks do your homework first--use the search function.
Ok so that 47 wpm...is that trained or untrained? According to the second poster you need to be trained to know how to 'chord.' However people don't need to be trained to type. I had typing classes within other classes and now I type at near 200 wpm with 100 percent accuracy on a good day. At worst I get about 100 wpm. Twice as fast as those 'chordists.'
200 wpm? bullshit! According to the Guinness Book of World Records (as of 2005), the fastest typist in the world is clocked at 212 wpm in short bursts and an average of 150 wpm over a 50 minute period. All this on a Dvorak keyboard, which likely gives a boost of 20%. You would be an extremely good typist to sustain 100 wpm, even on a Dvorak keyboard.
That leads me to another point: input to a computer is not going to change significantly any time soon. We have been using the QWERTY keyboard for 150 years, in spite of the fact that the Dvorak layout is "better", and is, itself, 75 years old. The reason is obvious: no one could be bothered to re-train data entry. No, the QWERTY keyboard will continue to be used until we have a device that is 100% perfect at speech recognition
I also agree that the analogue keyboard as we know it is on it's last legs. A lot of Mac users already use several input devices.. keyboard, mouse, Wacom tablet, specific Audio/Video editing keyboards. How low-tech is that! I love my Wacom tablet and use it almost as much as my mouse, however I would jump at the idea of having one multitouch device which can be a keyboard, Wacom, video editor... all controlled by software. This really is where multitouch comes into it's own. Not just in displays, but as a highly configurable input method.
If I could buy one tomorrow for anything up to £500 ($900USD) I'd would.
There was a lot of discussion and speculation on Multi-Touch, chording, gestures, virtual keyboards and the now famous Jeff Han video cited in the previuos post by turnwrite. Kolchak is the resident forum expert since he's a long-time user of Multi-Touch via the old FingerWorks iGestures pad. You can see his posts in the previously mentioned threads.
I wouldn't say I'm an expert, just somebody who's used it for a while.
One thing about some of these so-called "studies" is that they claim that subjects have gone up to about 50wpm. But what were the subjects' original speed touchtyping? Moreover, other studies say that chording takes a lot of training and that memory of less-frequently-used chords decays fairly quickly. Which isn't that surprising. Most people can't remember where certain menu items are if they don't use them often, so how can they be expected to remember chords?
Here's an interesting keyboard, the Alphagrip. I'm not sure I buy that the inventor was typing at 50wpm, but it seems interesting enough. Not enough for me to risk $100 on it, though. For that kind of money, I'd sooner buy a Comfort Ergonomic Keyboard.
200 wpm? bullshit! According to the Guinness Book of World Records (as of 2005), the fastest typist in the world is clocked at 212 wpm in short bursts and an average of 150 wpm over a 50 minute period. All this on a Dvorak keyboard, which likely gives a boost of 20%. You would be an extremely good typist to sustain 100 wpm, even on a Dvorak keyboard.
That leads me to another point: input to a computer is not going to change significantly any time soon. We have been using the QWERTY keyboard for 150 years, in spite of the fact that the Dvorak layout is "better", and is, itself, 75 years old. The reason is obvious: no one could be bothered to re-train data entry. No, the QWERTY keyboard will continue to be used until we have a device that is 100% perfect at speech recognition
Well maybe then UltraKey is bs. That's just what the program said...I thought it was too fast too, but I guess I trust computers too much lol.
But 100 words is about natural for me to do. Especially when I'm copying something or have something planned out. When I'm typing and thinking I'm slower.
2. I predict we are going from keyboards to keyboard + speech recognition. The keyboard won't be totally gone because it would suck to use speech recognition in several situations, such as a loud place, a place where quiet is required, a place where several people would need to speak at once, etc.
People keep saying "the keyboard is old and should go" but a fact is that every alternative out there doesn't stack up to the good ol' keyboard.
They keyboard ain't dead. Try programming with a multitouch device that doesn't give you feedback that you can feel (pressing a real physical button).
People keep saying "the keyboard is old and should go" but a fact is that every alternative out there doesn't stack up to the good ol' keyboard.
I wish I had the time to do a mockup video of what I'm talking about, but suffice it to say that in the system I'm talking about there is tactile feedback.
Just put your hands on the keyboard and then slowly type a "Y". If you aren't a hunt-and-pecker then you should have your pointer finger outstreched with the remaining fingers still closely grouped together. Now, holding that configuration tap all your fingers on your desk.
Did you feel your fingers touch the desk? Are you aware of how your fingers are orientated? Then you have your feedback.
Yes, this can be faster than a keyboard, especially when coupled with the other non-keyboard interactions that we use (mouse type functions) and gesture inputs.
Rich
With a multitouch device, I could see a virtual keyboard for backup or special situations. One could also just open up a window for finger drawing the text too. I also would like to see a stylus for fine detail work. Just as an artist picks up a pencil or paint brush for fine detail work, so too should a computer artist.
Thus computer input would be much more inline with real life. We'd use speech for most text entry, a virtual keyboard or handwriting pad for backup, use our fingers and multitouch for selecting tabs, buttons, scrolling, and general mouse input, and use a stylus for detail work.
Of course, one could always hook up a traditional keyboard and mouse so people will have options and choose whatever method they want to use. I forsee a transition period over a decade maybe.
I have a hard time seeing anybody reasonable not agreeing that this is the future. The only question mark is whether speech recognition is good enough. At some point it will be. Is that time now? If so, Apple could come out very soon with a new line of computers that implement this user interface revolution.
My guess is that it will be the operating system after OS X. However, there is nothing stopping Apple from jumping directly from OS 10.5 to OS 11. I doubt Apple is going to continue incrementing OS X all the way up to OS 10.99 and then release their new OS. In fact Apple may not even call this new computer a Mac. It will be so different that they might introduce a whole new brand to help people grasp the technology shift.
I really think Apple could introduce this new computer and OS at any time but probably won't for a year or two. We know Leapord is coming out and I doubt Apple would pull the rug out from underneath that work. Plus, a new OS is worthless without apps.
But this is the kicker. Apple makes so many applications now that they could port all their apps over to the new OS and UI. Thus when it comes out, they have a full suite of all the needed applications. This is the biggest hurdle for a new OS. Yet Apple can overcome this.
In another thread on this site, I bring up the OpenDoc concept too for the new OS. We have CoreAudio, CoreVideo, CoreAnimation - how about CoreText, CoreSpreadsheet, CoreWeb, and many other Cores? Also by incorporating media library apps into the OS - like iPhoto and iTunes - one comes up with a revolutionary OS that is compelling and needed.
New UI and input methodologies, new form factor, new OS and apps - I think the future is quickly coming into focus. I hope Apple has their employees very busy pursuing this vision. If so, they would need to keep this secret big time but I have a feeling that would be pretty hard to accomplish. The fact we don't have a torrent of leaks surprises me.
On the other hand, most of these pieces of the puzzle fit into the paradigm today so most of it could be developed right under our noses and we would not see it. Only in the last stages do things need to port over to the new OS.
Speech recognition is the future of text input. This is the most natural way for us to interact. When speech recognition gets good enough, most of us will use a keyboard very little.
You are joking right?
- Typing is much faster than speech.
- A regular office will then be like this: people babbling to the computer all the time. Very annoying (silence dammit!) but will also confuse all the individual computer audio inputs (neighbour, be quiet, you are confusing my computer!).
Sorry, I don't think speech will ever be used like this. Perhaps to turn on my lights, or start my stereo, but that's all.You also forget that computers will need to understand the context in which we say things. You'll need some kind of Star Trek computer being able to communicate with you like a real human, which is nearly impossible.
Typing is much faster than speech.
A regular office will then be like this: people babbling to the computer all the time. Very annoying (silence dammit!) but will also confuse all the individual computer audio inputs (neighbour, be quiet, you are confusing my computer!).
Beg to differ. Most people can't type faster than about 50wpm, while they can usually talk more than twice as fast. Also, it wouldn't be that different from offices right now, where numerous people are often talking on phones at the same time. Decent noise-cancelling microphones or headsets would suffice. The only exception is in numeric data entry, where keyboard data doesn't consist of entire words and can be entered about as fast as speech.
The only real problem is that voice recognition still has an error rate that's too high, mainly due to contextual misunderstandings.
If a person does need to improve their speaking ability, then the computer will help them improve this skill. This is not a bad thing and it helps not only interacting with computers, but helps interacting with other people.
Sometimes people would perfer to type and that is okay. Sometimes people want to enter confidential information. So there will always be keyboard entry. But most people would rather speak to a computer if they could. I know I would.
As for noise level rising in the office do to speeking - true. But this is not that different from people talking in the office now, not to mention fan noise, light hum, NVAC and other noises. In fact, people talk at a restaurant in small groups or some other social place without problem.
As for a computer being able to separate the user voice from the background, this is easy. All audio engineers know this is possible with cardioid miking at close range with a noise gate. A simple lightweight headset mic would solve the noisy office problem.
As for speech-to-text being too complex for a computer to handle, I disagree. Yes, it is hard, but not impossible. Computers keep getting more memory, faster CPUs, more CPUs, more hard drive space, better software. At some point, the hardware and software will be good enough. The question is when this threshold will be crossed.
Thinking that speech-to-text combined with multitouch is not the future of computer interfaces is foolish. It is just like the when the GUI was introduced to replace the command line interface. Command line interface users said their typing was faster then GUIs. Many people, especially geeks scoffed at the Mac. Yet look what happened.
Now, new input methods are being introduced and somebody someday will introduce a workable computer with speech-to-text and multitouch. Some people will knock it just like they did with the GUI on the first Macs. Yet, I think this paradigm shift will not be stopped.
I really think Apple will be the company to introduce this. If it can be done, they will do it. This is where they wanted to go 20 years ago. It made sense then and it makes sense now. The only hinderance has been the hardware and software needed.
Eventually the new interface and input mechanisms will replace the old. But just like we still have Terminal and some command line situations, so too will we have keyboard, mouse, or other inputs. The old ways will not be forgotten but I think they will be replaced by the majority of users in the majority of situations.
When I first started programming 20 years ago I had a computer with less than 64k of RAM. Now my computer has 4 GB and over 7 TB of HD space. That first computer ran at 4 MHz. Now I have dual 2.5 GHz. My first screen was about 480 x 320. Now my 30" LCD is 2560 x 1600.
Think what computer power we will have 20 years from now. At some point things not possible yesterday will be possible tomorrow. Maybe even today. I really think we are not that far off from a big shift in technology. I don't think an OS X version will be that shift but I do think OS 11 will be it.
My question is, is there going to be an OS.10.6 and OS 10.7, ... OS 10.9? What are these versions going to introduce to make me upgrade? What comes next? Apple spend two years developing the iPhone. That means for two years they have been asking the question what to do with multitouch on a regular computer.
Speech recognition is the future of text input. This is the most natural way for us to interact. When speech recognition gets good enough, most of us will use a keyboard very little.
This technology is already here, but it doesn't really work well enough to use in every day life. Haven't you called your cell phone company or the bank and spoken to the computer. They attempt to have you say simple words but you still have to repeat your request several times before it understands what you've said.
There are too many accents and variations on the way people talk for a computer to possibly understand them all clearly enough to write what is being said. What if you have a stutter or a lisp. If this could work reliably we would have it now.