While thinking about a topic for this post, I listened to a podcast I had not previously included in my online playlist (https://www.til-technology.com/my-playlist). I have just corrected this oversight, along with making another minor update. The podcast in question is CYBER, produced by Vice. This podcast is roughly weekly, and usually focuses on InfoSec-related news and issues, including interviews with some of the reporters who break the news.
In this particular case, however, the story that caught my attention was not strictly InfoSec. Instead, it was talking about something I had not previously heard about - “CharaChorder”, which host Matthew Gault covered in an article (https://www.vice.com/en/article/3abavv/this-keyboard-lets-people-type-so-fast-its-banned-from-typing-competitions).
Now, I love my mechanical keyboard (https://www.til-technology.com/post/it-s-easy-to-get-a-good-keyboard-right-er-right), but the speeds being described on the podcast and in the article certainly caught my attention.
When I was researching keyboards, I read a bit about typing speeds and keyboard layouts and such. For reference/comparison, I just tried two passes of the 3-minute test at https://www.typing.com, and scored 56wpm (words per minute) at 98% accuracy, and 58wpm at 99% accuracy.
Now, transcribing existing text is somewhat different from thinking about it on the fly, so I suspect my effective rate while actually typing this post would be somewhat better than that, as I find that I am often able to keep up with what I am thinking about reasonably well. Also, I find it somewhat distracting the way the site handles errors, and my standard practice is to add two spaces at the end of a sentence.
Excuses aside, however, I’m reasonably confident that my typing speed averages above 60wpm.
This is why Matthew Gault and Jason Koebler caught my attention by describing the CharaChorder. Jason Koebler described himself as typing at about 80wpm, and the fastest typists in the world as typing as fast as 200wpm. The founder of CharaChorder was then described as being able to type at 520wpm.
Aside from being interested in keyboards, I find it interesting when people come up with new ways to do things. This was quite exciting, but my BS detector started flashing, so I wanted to poke at this a bit more. This led me to several articles, several videos, and a look at the product page.
So, is it BS?
From my brief investigation, I don’t think so. The company appears to be at the early startup stage, though – it might survive, but maybe not. If I had more time on my hands, I might give thought to buying one and playing with it, but will most likely keep an eye on the company and the technology for a while, to see whether it starts to take off.
How can it possibly be so much faster than a traditional keyboard? To understand this, consider what a keyboard is (https://en.wikipedia.org/wiki/Computer_keyboard). A “traditional” keyboard has a single key (or combination, for example using the Shift key) for each letter/symbol. This means that the achievable speed is dependent on the user’s ability to move their fingers from key to key with sufficient force to trigger the desired key, which means that there are physical limitations to the speed of entry.
The CharaChorder addresses this by using multi-directional joysticks rather than single-directional switches. Consider the potential increase in speed if you could generate all of the letters in the alphabet without having to move your fingers off the “home row” on your typewriter. The amount of motion needed would be decreased dramatically, which would allow for greater potential speed.
While helpful, this is only part of the solution being put forward. Another component is referred to as “chording”. As I understand it, this is the way court stenograph machines work (https://en.wikipedia.org/wiki/Stenotype). Rather than “typing” single letters, the stenographer presses multiple keys simultaneously to represent syllables or words. Now, the “raw” output of a stenograph machine is gibberish to most people, but modern systems automatically convert it to English.
With CharaChorder, you can define “chords” and simply “mash” the appropriate keys (ie, press them simultaneously) to generate the word, rather than having to type each letter. It’s the combination of chording and the multi-directional switches that appears to generate the dramatic increase in the potential speed.
Interestingly, the company appears to be trying to raise funds for development of a traditional keyboard that supports chording. I suspect the idea is to improve typing speeds for people without requiring that they learn a whole new interface.
Which brings us to the “cheating” part. As Matthew Gault’s article mentioned, online typing competitions are flagging this new tool as a “cheat”, mainly because of speeds which exceed 300wpm. I think the inherent assumption for these competitions is that a “traditional” single-character-at-a-time keyboard is used. From this perspective, the multi-directional switches would be disallowed because they are not “traditional”, while chording would be disallowed because chords are not single-character-at-a-time.
I’m not interested in typing competitions, however. Like most people, I have work I want to do that involves entry of text, so if this new tool can provide me with a dramatic increase in speed without requiring years of training, I’m interested.
There are a few concerns, though. How long does this new approach take to learn? How long does it take to build proficiency? And how long does it take to get to the extraordinary speeds being advertised? Also, would this simply be a niche skill/tool, or could this sort of thing become common? All interesting questions – the vendor site suggests about a month of 30-60 minutes per day to achieve significant results. Overall, that doesn’t seem excessive for such dramatic potential results – I’ve read about people switching to Dvorak or another keyboard layout and spending comparable time to gain relatively insignificant benefit.
One other thing, though. The chording feature appears to be designed such that it could allow the setup of what are essentially macros to insert pre-defined blocks of text. For example, if I like to end an email with “Thank you and have a wonderful day! RG”, I could save a significant amount of time if I could include that entire block of text with a single chord.
An interesting rabbit-hole indeed! I plan to keep an eye on this, and may post again – unless the telepathic interface I asked a certain developer for about twenty years ago comes first... I guess we’ll see.
Cheers!
Comments