Dragon Professional Individual 15

Dragon from video game Skyrim

No need to shout, the Dragon understands my Nordic dialect right away!

Over the years, I have made a habit of reviewing the various versions of Dragon NaturallySpeaking. Lately, Nuance has stopped using the phrase NaturallySpeaking in most contexts, but it is still the same product, and it is now up to version 15.

As the software has become more expensive again, and as it is already good enough for my limited use, I have started skipping some versions. Dragon version 13 was already good enough that I did not really expect it to get any better. Impressively, Dragon version 15 is actually noticeably better right out of the box.

Dragon version 15 uses a new “deep learning” technology similar to what is used in the most successful artificial intelligence projects. Dragon has always (or at least for as long as I have used it) had the ability to improve based on feedback from the user, as well as adapt its vocabulary and writing style by reading through documents. While these options still exist, there is less focus on them now as Dragon quietly adjusts in the background during everyday use.

Dragon has also clearly had some opportunity to acquaint itself with human speech in general before shipping to the customer: The product is amazingly accurate right out of the box. Longtime readers (if any) may remember that I compared some of the early versions to homesick exchange students from other continents. That time is long gone. Dragon version 15 understands even my “Skyrim” pronunciation of English (I grew up in Norway in the 1960s, where even the English teachers has rarely if ever been to England, let alone America or Australia.)

There is one problem that has dogged this software from the start, and it still remains, even if just barely. When we speak, we don’t actually pronounce periods at the end of the sentence; rather, we slightly change the tone of our pronunciation toward the end, typically speaking less forcefully. Conversely, we don’t actually pronounce a capital character at the beginning of a sentence; instead, we pronounce the first sound slightly differently from the rest. Ideally, speech recognition software might be able to use this to take dictation without requiring us to specify punctuation. Dragon NaturallySpeaking used to have this functionality, but I gave up on it pretty quickly. What actually happens is that even when I dictate punctuation, there is a slight increase in mistakes at the very beginning and end of the sentences. This is especially true if I don’t pronounce some form of punctuation at the end of my string of words, for instance because I run out of breath during a long sentence. I have to say, however, that this problem has been almost eradicated in the latest version of Dragon.

To me, recognition accuracy is by far the most important part of any speech recognition engine. But Dragon 15 has also some other features in addition to the improved accuracy. It has better support for various modern software, and it allows voice activated macros. (I believe this feature was also in version 13, but I did not use it then and I don’t use it now. In any case, functions like “insert signature” should be part of your email software, rather than your speech recognition software.) Also, the big unnecessarily helpful sidebar with examples no longer starts up by default. It used to do, and is also used to permanently displace any windows that happened to be in its way.

As usual, I am including a paragraph where I don’t in any way correct this transcription. This is that paragraph. (It may not be obvious to the reader, but that should be “the transcription” in the first line above.) Dragon used to be available in a few languages besides English; I am pretty sure I saw touch at some point, and Japanese? I can’t find any trace of that now, but I will admit that I have not looked very carefully.

Not too bad, huh? That should of course not be “touch” in the previous paragraph, but rather Dutch, the language in the Netherlands. (It actually got it right this time without correction. Go figure.)

MS Windows troubles

Screenshot anime Kanojo ga Flag o Oraretara

This morning was absolutely crawling with chaos. It started as I turned on my home office computer, which had installed updates at 3AM and restarted itself, as it frequently does. It seems like a good idea, to install updates while you sleep. After all, you would not want to miss the latest security patches and improved functionality.

Unfortunately, the new functionality was that I could not log in. Whether I picked my usual account or the betatester account I use for testing games, there was just a brief pause and then Windows returned me to the login screen. No error message. I restarted the computer and tried again. I did various things and tried again and again. No change. I restarted in Safe Mode. Same problem. I restored Windows to last good configuration. Still the same.

I installed Ubuntu Linux, which is a pretty good alternative to Windows for most people, and free. After a little while I switched to Xubuntu (it is really just a different setup, the core is the same as Ubuntu, but Xubuntu is more similar to old Windows versions). Ubuntu is free, like most Linux versions. I use to install it on old laptops when they become too slow under Windows. This is less of a problem these days, but it was a big deal back in the days of Windows Vista.

Xubuntu is nice enough, but there were a couple problems. I had used this machine to provide Internet access to my cabled home network, which includes a Windows 10 machine for playing games, a NAS (home server) for backup and sharing files, and a small old notebook computer for uploading and downloading to and from the NAT without taking up resources on the main machines. But now I could not get Linux to share the Internet. It should be easy, really, there is a choice for it. “Shared with other computers” it says, but that actually only lasted for a minute or so, then I got a message “Disconnected from Ethernet”. (Ethernet is the cabled network, to put it simply.) I did various things and restarted numerous times to no avail.

Eventually I found an USB wireless receiver and connected this to the Windows 10 machine, then told it to share its Internet. This worked well enough, except the NAS (Network-Attached Storage) server did not show up. After changing the workgroup name by editing a configuration file, I got it to show up. But as soon as I tried to copy a file to it, it hung up and show up empty until I logged off an logged on again. This repeated itself for as long as I bothered trying.

I was kind of in a hurry to continue working on my National Novel Writing Month story. Luckily that was saved on a disk I could access from Xubuntu. I copied it to a USB drive, in case I wanted to continue writing on it on the other Windows computer (the gaming computer). I installed WINE, a program that lets you run Windows programs in Linux. I had already read a few years ago that you could run yWriter in Linux this way. (yWriter is the program I use for writing novels. It is written by a programmer and novelist and fits my working style exactly.) It did work when started with WINE, and it found my novel in progress, but the spell check did not work and it did not recognize the names and locations. I downloaded the dictionary and manually copied it to the place it should be. Now it worked except it did not recognize words when Capitalized, such as at the start of every sentence.

Somewhere around this time I decided to reinstall Windows on one of the disks. (I am keeping Xubuntu on the other.) This took the rest of the evening and will continue into the next day or two or more.

Needless to say, there was no progress on the novel this day. But then again, contrary to the slogan of National Novel Writing Month, the world does not really need my novel. Probably.

The Dragon Upgraded

Screenshot anime YowaPeda

I feel like I can go anywhere… With Dragon NaturallySpeaking Premium, you can dictate anywhere using a USB microphone, wireless microphone, smart phone or dictation device. (Even so, I don’t recommend dictating while biking!)

In the past, Dragon NaturallySpeaking has been available in several different versions, and I have always used the cheapest one, Dragon NaturallySpeaking Home. It usually cost around $100, with the occasional big sale where you might buy it to at half price. As an existing customer, I could also upgrade it to the next version at half price from the start. Last time, two years ago, I also did that; I even preordered it.

This time, there was no question of pre-ordering. Either they didn’t ask me, or I missed it somehow. My first hint that there was a new version available came from a mail that offered to let me upgrade to Dragon NaturallySpeaking 13 Premium for €99. A bit more, but then the Premium version has some unnecessary but nifty features. So instead of being my usual cheapskate, I went for the premium version this time. It was already available for download; there was a link in the mail to the website where I could buy it. I checked the requirements and looked for any traps, but that didn’t seem to be anything suspicious. So I bought it with credit card, and could immediately start downloading.

The installation was easy and trouble-free, although it took some time. I first downloaded a small installer program, which then downloaded the big installer program, which then unpacked to a separate folder, which then installed the program in the default location. It may sound a bit complicated, but it was mostly just pressing the “next” button, although I had to choose a directory for the temporary files. I saved them to the network drive in case I have to reinstall on this machine or another. I would also recommend using an external disk for the temporary folder if you have limited disk space, since at some point there will be three big files and folders taking up space simultaneously: The big installer, the folder with the unpacked files, and the actual installation in your Program Files folder. If you have a reasonably new computer, this would probably not be a problem.

Speaking of new and old computers, the two latest versions have each reduced the computing requirements, so that you can actually run version 13 faster on a weaker computer than version 11. Good work!

After installation, Dragon offered to upgrade existing user profiles. This took surprisingly long, even for the profile that was almost empty. Several times I wondered if it had crashed, but I didn’t need to use it immediately so I just checked in on it from time to time, and eventually completed. If you don’t have an earlier version of Dragon, each with an effort to create a new user profile instead. I believe that in this case, you will also be offered to train the program to recognize your voice and improve its accuracy. At least this happened in the earlier versions. It may be that it is so good right out of the box that they don’t bother with that now?

As far as I remember, version 12 looked very much like version 11. Version 13 has a whole new visual profile, so it is obvious at a glance that you are running the new version. The DragonBar, usually placed at the top of the screen, is now just a small button when not in use. If you move your mouse to it, it expands to become larger than it was in the previous version, and the microphone on/off button also becomes much larger. The “Learning Center” (formerly Dragon Sidebar) still takes up the margin of the screen, but it now has a black and white color scheme and also seems to have larger letters. As always, you can minimize or remove this Learning Center if you don’t want (context dependent) hints about what you can do next. Even the DragonBar itself can be minimized to the system tray, and you can access the most common functions by shortcut keys or by voice commands. But that has always been the case, I just wanted to mention it.

As I mentioned in the previous entry, the first thing I noticed when trying Dragon NaturallySpeaking 13 was the leap in accuracy. I realize that I have praised its leaps in accuracy since at least version 10, but this time the difference seemed to me bigger than the official count of 15% improvement. 15% improvement does not seem a lot when the accuracy is already claimed to be about 99%. To me, it seems more like it has increased from 99% to 99.5%, which would actually be a doubling of the accuracy in the sense that there would be half as many errors. But I admit that in my case this could be because of an improvement in the handling of USB headsets.

(It is unfortunate that I cannot maintain this level of accuracy for longer texts, because my voice becomes hoarse after a few minutes. But this is an affliction that I share with very few humans. One hypothesis is that it comes from my years of almost complete silence, where I only asked a few questions at work and did not speak at all on my free time. If I take breaks and drink a little water between the paragraphs, I can continue for longer.)

The premium version contains some features not found in the home version. For instance, you can now make the program read back your own voice, not just a synthetic text to speech rendition of the text. You can also use a smartphone as a microphone, or even use recording devices and have the program transcribe them later. There is supposedly also an option to create your own voice commands, basically macros, but I haven’t tested that yet.

In conclusion, Dragon NaturallySpeaking 13 is awesome. You can actually speak naturally to it, and with very little training it will put your words on the screen and let you control Windows. Upgrading from version 12 seems to make a big difference for me, but your mileage may vary. Upgrading from Home to Premium is probably not a priority unless you have a USB microphone or some other unorthodox input device, but it adds some fun new features.

(As usual when writing about dictation software, I have dictated this entry in its entirety, except for a few minor corrections.)

Dragon NaturallySpeaking 13

Squeeing girls from anime Gekkan Shoujo Nozaki-kun

This is how I think my readers should react when I write about Dragon NaturallySpeaking speech recognition from Nuance. Somehow that never seems to happen. Let me try again, it’s two years since last time.

I love living in the future. And one of the more futuristic things that I have is the speech recognition software for Windows, Dragon NaturallySpeaking. (Windows also has its own built-in speech recognition, but for those who can afford it, Dragon is definitely the one hardest to distinguish from magic.)

Today I got a mail from Nuance, offering to upgrade my Dragon NaturallySpeaking 12 Home to Dragon NaturallySpeaking 13 Premium for €99. I immediately grabbed the chance, just as I have done every time there was an upgrade for the last five years at least. Was it worth it? Well, to paraphrase a friend of mine, €99 is a lot of money if you don’t have it. This is obviously not a product for the working classes of the developing world, but for Norwegian office worker the amount is trivial, barely noticeable against the high salaries and the high prices up here. And for me at least the effect of the upgrade was dramatic.

According to their website, version 13 is 15% more accurate than version 12 right out of the box. Evidently this has either crossed some kind of threshold in my case, or there was some bug in the version 12 Home in relation to my Plantronics USB headset. The USB headset worked very poorly with the previous version on my laptop (although it had worked reasonably well on the desktop with version 11). So when I wanted to dictate, I had to take off my USB headset and put on an analog headset for the duration, and even then the accuracy was at most marginally better than in version 11. Today after the upgrade, I can use my USB headset again, and what’s more: The accuracy is more than 99%. It still makes mistakes, but less so than my fingers. (And I have been typing for almost 50 years now.)

Back when I wrote about an early version of Dragon NaturallySpeaking here in the Chaos Node, it had only entertainment value for me, although I realized it could be useful for people who could no longer use their arms at all. Some years later a newer version helped save me from disability when my job caused a serious case of repetitive strain injury. At that time it still made quite a few mistakes, but at least I could correct them with my voice. Since then it has improved even more, and I have given it pretty good reviews each time. But let me tell you something: For me, version 13 is a giant leap.

It still makes mistakes, but so few mistakes that I risk overlooking them in the middle of all the perfect text. We are talking about perhaps one error for each paragraph on the first day. The software gets used to the sound of your voice and your writing style and also learns from all the errors you correct, so it gets better the more you use it. So to pull off this level of accuracy with almost no training is impressive indeed.

For those of you who are still here instead of being busy buying it, my next entry will get into some more detail about the installation and differences from the previous version.

(As usual when writing about dictation software, I have used the program to dictate this entry, except for a couple of minor corrections.)

Self-spaced repetition software

Screenshot Sims 3

Games are usually the enemy of studying, but some elements of games can be used to improve learning.

Spaced repetition is an amazing technique for learning without understanding. Understanding is certainly superior and in a league of its own, but it is hard to command or even predict; and sometimes you need to memorize for a while before you can understand. One may compare this to gathering ingredients before you can cook. Before you can read sentences, you must know the words, and so on.

Spaced repetition takes advantage of a particular memory effect: The best time to repeat something is just before you forget it. Repeating many times in a row adds little after the first couple repetitions. Repeating at fixed intervals helps, but the best effect comes with increased intervals. You may for instance double the intervals, which is more effective than fixed intervals. But the best effect is when one actively recalls a fact just as it is about to be forgotten: It should take a little effort to remember it, but one should be able to do it.

Since our brain is not under constant surveillance, the only person who can know this time is ourselves. Conventional Spaced Repetition Software (SRS) starts with a standard model of human memory, then adjusts intervals down if people keep forgetting, or up if they remember everything. Good SRS lets the user mark the difficulty with which they remember. But there is an even better solution, at least in some ways, and the Duolingo online language site has found it.

When you start a study unit in Duolingo, the “learning meter” is empty. As you learn, it fills up until it is full. But as time passes afterwards, the learning meter begins to slide down toward empty again. You can see this for each topic, and for each “workbook” in a topic. (This is a session that typically takes 10-15 minutes to go through.) You can then go back and test yourself.

If you go back and run the test while you remember everything, you will certainly get the learning meter back to full again, but the experience will be rather boring. Duolingo is a very game-like learning system, where you have to translate back and forth, listen to sentences in the target language, describe pictures or pick from multiple choices. You never know what the next question will be. But if you know everything by heart, it is not very exciting. It is like winning chess against a small child.

On the other hand, if you wait too long and have forgotten the words or phrases, you will be thrown out after three mistakes and have to start that workbook over until you get it right. That is not too much fun either, even though it only costs you a few minutes.

The result is that the users themselves find out how long to wait to get the most rewarding “game” experience. If it was too easy, you learn to wait longer. If it was too hard, you make sure to return earlier next time. There is no need for the software to know whether you learn fast or slowly. All it needs to do is reward you when you get the balance right, and it does so with an exciting learning experience and a feeling of winning against a worthy opponent: Your own forgetfulness. You pick your battles, so with a little experience you pick the best time yourself.

I am not sure how easy this is to translate into other forms of learning, but I think it may be easy with anything that requires memorization of facts. Geography and history come to mind. Make small, focused units and a table where one can see which ones begin to slip. Adjust the speed at which they slide depending on past performance. It can probably be done better than it is in DuoLingo, but the principle works amazingly well. Humans are very good at learning things when having fun.

Memrise vs Anki: place yer bets

Instead of doubling the amount of time I spend studying, I am trying to double the precision. Although you can learn almost anything by repeating it 7×70 times, the best time is just as you are about to forget. Anything before or after is less effective. But how do you know when you are forgetting if you don’t remember it?

I have written quite a bit over the past month about Memrise, a free Web resource for memorizing facts, vocabularies etc. It combines two of the most powerful techniques for rote learning: Spaced repetition and mnemonics. Spaced repetition tries to make you recall the fact just before you forget it, as this causes maximum learning with minimum effort. Mnemonics try to associate random facts with something that is easier to remember. This is obviously most effective if you do it yourself, but that can be frustrating. Memrise uses associations volunteered by users, and you can add your own.

I rather like this approach, and the way you can study at your own pace. Unfortunately, most of the time I remember 66% at best rather than the 90% that is the goal of spaced repetition. This was also the problem with the two previous SRS programs I used, AnyMemo and Mnemosyne. (Spaced Repetition Software is SRS business!) So I am testing another free program, ANKI, which has a good reputation among self-study amateur linguists. I am not too optimistic though. Now that this is my fourth attempt, I may have to accept that it is I who am too old for the programs that fit most people. It is the same with physical exercise, after all, but there I can set my own pace. And that’s the thing.

What I really miss is a dial or lever I can set, so the software reminds after e.g. 90% of the time it thinks should be right. Clearly the programs all overestimate my memory for random words. Of course, it would probably have helped if it was not so random, if it was at least somewhat related to my ordinary life. But that’s not what I need it for. I would really like something that was adjustable to me, rather than the other way around. It is kind of discouraging to have forgotten a third or more of the words when it is time to review them. It is also bad for learning – the “memory traces” in the brain weaken more quickly after the ideal recall time, or so I’ve read. So ironically, I would probably even spend less time reviewing if I had that “confidence dial”.

Anki does not have that, but it does have levels in the answers. Instead of just checking for itself whether you got it right, it asks whether it was hard, good or easy. The ideal is good, which is when you remember it with a little effort. If you had to think long and hard, it goes easier on you with that word or fact next time, in the form of asking you earlier. If you say it was too easy, it waits longer. And if you don’t get it right at all, it shows it again very quickly. So that sounds like an improvement.

On the other hand, I liked the suggestions for memorizing words, and I liked the way Memrise used different forms of multiple-choice questions in the early phase of learning a new word, then giving more and more options and eventually requiring you to write the answer. It also requires writing when reviewing, which involves more of the brain and makes it harder to fool yourself (“well, I got it ALMOST right!”).

I have picked up Anki and installed it on my PC and my Galaxy Note 2. (Unlike Memrise which is a website but requires some advanced browser features and can’t be used on my mobile devices.) Anki is also easily synchronized between two (or even more) devices. There are a lot of premade vocabularies and other data sets, and it pleases me to see that a lot of them are for studying Japanese. I downloaded a fairly small one that is mostly tangential to what I have already learned, and am testing it now.

Unfortunately there are obvious errors in the dataset I am testing, although small ones. Occasionally a romaji (western character) is used in a word written with katakana. I saw one obvious misspelling beyond that already in Japanese, and another in the English text. The Japanese is written in a font that is like an uglier Japanese version of Comic Sans. I hope this is a feature of that particular set and not of Anki! It is quite hard to read after the very legible font on my Windows machine, not to mention the downright beautiful hiragana font on the PC running Ubuntu Linux.

Apart from that, it seems nice enough. With the mobile app I can study at the bus, during breaks at work, even while a game is loading. OK, not much since I have a fast machine. But still, very handy. And I like its approach: If I don’t recognize a word, Anki shows it again after a minute. Once I recognize it, it increases to 10 minutes, then a day. I inserted 1 hour between those, the system lets you add steps like that. Then it goes up to 4 days and so on, I am not sure how far it goes. The most important part is of course whether I actually learn the words. I will have to come back to that. But if it turns out to wait too long, like all the rest, I will try to choose “hard” instead of “good” even when I remember, and see if that fixes it.

I really hope I won’t have to write my own. There are already quite a number of these. There’s Supermemo, the original and possibly best, if you can live with complicated. And there’s at least one other that I forgot the name of. I do that a lot, forget names. Although I don’t always remember doing it.

Talk to your toaster

I also used to be excited about the future, but now that I live here, I take it for granted.

NaNoWriMo – national novel writing month – is approaching once again. (“The month formerly known as November”, as I like to call it.) The forums for 2012 are up and running, and in the technology section there is as usually a thread dedicated to speech recognition, or more specifically Dragon NaturallySpeaking. (I would not mind a more general thread, since Windows also comes with speech recognition built in. Hopefully we can have more threads later.)

One thing I wanted to say early on was that it is not enough to be able to use speech recognition in a technical sense. The next challenge is to be able to tell a story to the computer. This is a very different thing, especially for us who have been writers for many years and are used to thinking with our fingers. It also doesn’t help to have been a grown-up for many years, during which you have not been able to tell long, obviously made-up stories to people without them looking at you very strangely. I suppose there are some families in which this problem does not exist, but I am not sure whether it is a good thing or not… ^_^;

So I recommended that people start telling stories to their computers already now, all through October, so that they have gotten over that hurdle, that shyness or awkwardness of telling imaginary stories out loud to inanimate objects. In fact, I recommend practicing on the toaster as well, and with blatant nonsense. The purpose is not to deliver the Great American Novel to your amazed toaster, but to get yourself to accept the unreasonable fact that it is possible to tell stories to home equipment. Such are the times in which we live. I could not have made it up in a sci-fi novel. Magic fantasy, perhaps, just perhaps.

I ask you, gentle reader, to consider this: Not only do I occasionally talk to a machine without being insane (or more so than those who don’t). I also carry in my shirt pocket a telephone, my own library with dozens of books, a bookstore with millions more, thousands of newspapers from all over the world, millions of songs and an unknown number of movies, and enough cat pictures to last the craziest old cat lady for a lifetime.

You can probably add to this, but the point is: I do this almost every day without giving it a second thought. I don’t wake up each morning thinking: “Oh my God! I live in a miraculous, magical world filled with amazing wonders that I would not have believed were possible when I was a child – what should I do today to take advantage of this to the fullest?”

If I did, and if my conclusion was that I should start the day by talking to my kitchen equipment, that might not be the worst thing I have done in my life.

Chunks of memory

Your personality doesn’t matter. This is a skill you can learn.

Extreme feats of memory are possible when we remember large quantities of information as one unit, because we have spent so much time with it. We all do this.

I know I have written about this before, already back in the original Chaos Node, where I read about it in an issue of Scientific American. Recently I read about the same thing in the book Talent is overrated. For instance, chess grandmasters could glance at a chess board and later reconstruct it exactly, something no normal person could do. From another ancient board game, Go (or Igo), I know that high-level players not only remember every move of a match, but can often guess how a match has progressed even if they arrive late into it, possibly even at the end. For someone unfamiliar with the game, this seems like magic. And yet we all do the same thing.

Neurotypical humans store incredible amounts of information about the people around them. Not only can they recognize a friend at a glance after several years, they can also keep track of the relationships between everyone in a village: Who are second cousins with who, who are friends, who are enemies, who are in love with who and who were in love with their current enemies years ago. Nobody finds this remarkable in the least, but it is really amazing.

Likewise we are very good at categorizing things. Or dogegorizing, I guees: Even children can usually tell cats and dogs apart, although small children have trouble with really small dogs which may be labeled cats. Even though there are so many different dogs and breeds of dogs, people have this internal concept “dog” which kind sums up the essential elements of doghood and which they remember as a unit, even after many years.

In the same way, if you grew up with your mother, when you think “my mother” you actually remember thousands of things, from how she looked at various ages to details of her behavior and relationships. You don’t consciously think of all these details every time you see her name, but if someone were to say something untrue about her, you would recognize it immediately.

In other words, all of us have the ability to remember very complex things as 1 unit.

Our short term memory is very limited, usually we are able to remember around 7 units of information at the same time. The actual number may vary from 5 to 9 and can be increased with rigorous training. It is the number of digits you can remember while walking from one room to another without repeating them in your mind. But if those digits are familiar, the number suddenly increases dramatically. For instance, to me the 6 digits 271258 count as 1 unit of information: It is my date of birth in the format used in this part of the world, ddmmyy. So I would be able to remember 6 more digits while leaving the room. Yes, strange as this may seem, I have an average short-term memory. I have tested this.

***

The computer language Forth caught my attention toward the end of high school. It was little more than a rumor back then, some new-fangled invention from the States. Personal computers were something hobbyists built themselves, and pitifully weak. A corporate mainframe at that time was perhaps a match for a smartphone today. OK, perhaps a little more. Let’s say a smartphone next year. But only a few years later, I had my own personal computer, weak though it was, and was programming in Forth.

This particular computer language had a peculiar structure. The basic language was very simple, consisting of a stack for data and a set of basic “words” that were coded in machine language, either directly or with an assembler. These were very simple commands which would be defined differently from computer to computer because of the hardware, but which (ideally) had the same names and function on all computers. But this was not what fascinated me. Rather, you could define new words by combining the old ones. The new words could be used in the same way to combine into more words. By keeping the definitions short and simple, the risk of errors was greatly diminished, and it was easy to test the new words right away. Yet there was no obvious limit to what you could do. There was very little overhead in having many levels of definitions.

The reason this appealed to me is that I am a verbal person. I think in a very similar way to this computer language, building new concept from existing concept. As long as I keep it simple, I can trust the knowledge I build from basic, and I can test it.

***

When you spend a lot of time doing something, whether it is programming or chess or surgery, you acquire what is called “domain knowledge” within that area. And when this knowledge becomes a part of you, something as natural to you as cats and dogs and family and friends, you begin to be able to think of it in chunks. The chess player can remember every piece on the board because not only the pieces are familiar to him, but the possible configurations too. He has seen them many times: When this particular group of pieces appear on this part of the board, it means certain risks and opportunities that are very real to him. He has no need to memorize this particular picture: He has seen it before, repeatedly, and it has meaning to him.

When I learned to read, I had to learn the alphabet like people did for generations before me. I hear that this is no longer considered very important, people start looking at words as pictures right away. But words still consists of letters, and sentences consist of words, paragraphs of sentences and so on. When you remember a poem or a particularly moving passage from a book, you don’t try to recall each individual letter in turn. Like the programming language, the “primitives” – the basic components – soon become buried in higher-level structures. Reading and writing are themselves everyday examples of structured knowledge. And as with the programming language, there is no obvious upper limit. Scholars will hold entire books conceptually in their mind – not word by word probably, but still in a very real sense whole books – and compare them to arrive at a higher meaning from the way the books agree or disagree. If we were wiser and lived longer lives, who knows what we could achieve?

Humans, it seems to me, are not proportionate to the savanna or the shores from which the “naked ape” emerged, but rather proportionate to the infinite. As better men than I have noticed, the most incomprehensible thing about the universe may be that it is so comprehensible. At least now we know a little bit more about ourselves as well.

Dragon NaturallySpeaking 12 – part 2

“If you don’t listen to everything, you won’t understand anything.” When dictating, speak in statements, or at least phrases. Don’t stop randomly, for instance between “the” and noun.

I have now had the new version of Dragon NaturallySpeaking for a couple days. With my throat condition, that probably corresponds to a couple hours for those of you who talk a lot. I intend to use Dragon to dictate this entry, but I I will still need to make corrections. Perhaps you won’t, if you are a native English speaker without too much accent or dialect.

I am impressed by how quickly  Dragon has adapted to my voice.  It certainly happened much faster than with any earlier version. In all fairness, I also have more experience with Dragon now. For instance, as I mentioned in my previous entry,  I have made sure to perform training at different times of the day and at the beginning and end of a “speech”.

(I actually dictated the previous paragraph without making any corrections, but that’s not the rule for longer paragraphs yet.)

*** 

A problem with browsers: I haven’t heard about this from anyone else, but I have found Dragon to operate erratically in text entry fields in browsers. This could be a serious drawback, considering how much time we spend on the Internet these days, both at home and in the office. At first I thought the problem was only with Opera, which is my browser of choice. This program is not explicitly supported by Dragon, and in version 11 the text field where I write my journal was marked as unrecognized. While I could try to dictate there, the result was usually pretty bad. In version 12, Dragon alternates between “unknown text field” and “normal mode”. If I dictate while in normal mode, it seems to work well enough. If it is in unknown mode, I can usually just wait and it will switch to normal mode  after a few seconds. Even so, the hotkeys don’t work, and corrections  frequently mess up the text a little. So for longer texts,  I tend to use the DragonPad and just paste the result into the browser.

Unfortunately, I have similar problems in Internet Explorer when using Google+. Again, this may be a problem with that particular application – even typing can sometimes be sluggish in Google+ – but there are tens of millions of people who use that application frequently. Then again, it might be just me. Since I am one of the first to actually buy the product, there aren’t much in the way of reviews for me to compare with.

Is this a big deal? After a few days, you would probably not need to make corrections every time you post. A more serious problem might be if parts of the text are missing because you dictated while it was in “unknown field” mode. Again, this could be peculiar to my computer – there certainly doesn’t seem to be any problems in the demonstrations on YouTube. (Then again, they use neither Opera nor Google plus.)

***

 I haven’t had any problems with other programs. Dragon works beautifully with yWriter, the program I use when writing fiction. It seems to work fine with all kinds of notepads, whether plaintext or rich text. The commands for opening programs, switching between programs or clicking on buttons work as expected. And the on-screen help which came with version 11 makes it unnecessary to memorize the handbook with its dozens and dozens of commands. I am sure there are a number of features that I am never going to use, but better that than the other way around. And in version 12 you can even turn off features at a very detailed level if you’re afraid of activating them by mistake or if you simply need more speed.

You guys, I really feel like I can’t get across how smart this program is. When I first tried Dragon NaturallySpeaking approximately a decade ago, I compared it to a drunk and homesick high school exchange student. I compared version 11 to a native English speaker with a college education. But version 12… It is like a professional secretary with a genius IQ. Oh, it still has problems now and then, but it has only spent a couple hours with me, and there are several sounds in English that Scandinavians of my generation simply cannot pronounce. I am not sure any of my English-speaking readers would be able to understand me that well after listening to me for a couple hours.

Because I have spent decades mostly in silence, I cannot dictate a long entry like this without taking breaks. My voice simply dries up. If not for this physical handicap, I would be sorely tempted to do exactly what Nuance proposes in its slogan: “Stop typing, start talking.” It really is that impressive.

The Dragon has landed!

 

Dragon NaturallySpeaking 12 became ready for download today for us  existing Dragon users who had pre-ordered. I’ll come back to the installation shortly.

For those who do not know, Dragon NaturallySpeaking is a voice input program for the Windows computer, and the leader in this category. It takes dictation but also allows you to open programs, search the web, compose mail and edit existing texts without using your hands. As such, it solves an acute problem for those who don’t have hands or can’t use them. For us who have hands, it is most useful for dictation. It is fast and, with a little practice, amazingly accurate. The new version claims a 20% increase in accuracy, putting it well above 99% accuracy with 15 minutes of training. In practice, it takes longer, but the program keeps learning the more you use it. When you see an experienced user work with Dragon 11.5 (the previous version) it is “indistinguishable from magic”.

Installation:  The download link from Nuance arrived by email before I woke up in the morning. A separate mail also contained link to the training video. While I am personally a fan of reading, the training video will surely be welcome by dyslexic users, another core customer group. (The program can also read text out loud, even text you have not dictated.)

The download process proceeds in several steps. You first download a tiny download manager program. It does not really matter much where you save this, it is very small. This program must be run to start the main download. The main download is a compressed file, but still close to 3 gigabytes. This must again be unpacked to a larger set of files before the actual installation. During the unpacking process, both the compressed file and the unpacked file take up space simultaneously, and that’s before the actual install into the Program Files directory. This program is not recommended for people with small disks!

It is recommended that you back up the compressed file so that you can install from this if your computer suddenly crashes or if you simply decide to buy a new at some point.

The download went without glitches, but the install itself caused me some trouble. A ways into the installation, the program warned me that several processes had to be closed down before it could continue. Three of these were unknown to me, and did not appear with the given names in Windows Task Manager. I had to break off the installation and reboot the computer, then run the install again. The install did not automatically resume, and if I had not taken note of where the unpacked file was saved, I would have had to restart from the compressed file. I would recommend you reboot your PC before you start downloading, and not start any unnecessary programs until after the install is complete.

After installation, the software offers to let you register the product online. There is also an online activation which is necessary to continue using the program. The registration and the activation are unrelated tasks.

As a user of version 11, I had my existing program removed automatically and my user account upgraded to the new version. This takes some time even on a fast computer. New users will be led through creating an account instead, and the system checks the quality of your microphone input before asking you to read a text to attune the program to your voice and reading rhythm. You can skip this step and train the program by correcting mistakes if you want. New users also get an offer to let the program read through their email and documents to adapt to their vocabulary. This is a separate task from adapting to your voice. Again, you can skip this and just train the program through use, if you are impatient, but there will be more errors during your first few days of use if so.

Accuracy training: Since Dragon was complaining about my microphone, I bought another, an analog headset to replace the digital USB headset. I established a new user account and started over from scratch with the new hardware. This microphone passes Dragon’s test with flying colors, but the new account doesn’t have any of the accumulated experience with my speaking. Newsflash: It certainly wasn’t useful right out of the box!

My experience is probably not typical, since I am a foreigner to the English language and also have a chronic problem with my vocal cords – my voice grows “rusty” many times faster than a normal human – but I think we should still consider this. After all, most people aren’t native English speakers, or if they are, they have dialects or accents. And your voice does change with use even if more slowly than mine. And my experience is that it takes several hours for a new user before Dragon NaturallySpeaking 12 becomes truly useful. So don’t buy this program an hour before you need it. Set aside a couple days at least to become good friends with it before you start working together.

Not only does your voice change after you have used it for a while, but it is also slightly different from morning to evening. So it may be a good idea to do some reading training at different times, to help the computer get familiar with your voice. It is not necessary to read all the way through the exercises, you can click finish at any time. Also, try to make sure that you read the exercises in the same way that you speak to the computer when you dictate. For my part, I have found that I have a tendency to speak faster and in longer stretches when I read something, compared to when I dictate my own thoughts. For some reason I also tend to read louder – perhaps a habit from my school days? We used to be required to read aloud in class.

Features: The previous version mostly improved the user interface, introducing context-sensitive help in the form of the “Dragon Sidebar”. It also expanded support for more programs, and the engine was made more efficient. Version 12 has very few changes in the user interface; it supposedly includes 100 new features, but I don’t expect to need more than a few of them. Most of the development this time seems to have concentrated on the technical: In addition to the improved accuracy, the program also runs much faster, especially on new computers where it now takes advantage of multicore processors and extra memory. Additionally, even the home version can now take advantage of mobile phones as microphones: If you have an iPhone or an Android smartphone and it’s on the same Wi-Fi network as your computer, you can dictate to your smartphone and have the text appear on your computer screen!

One feature I thought was included in the home version, but which evidently isn’t, is playback of your own dictation. On the other hand, the program includes an excellent synthetic voice which can read what you have dictated (or any other normal text). This will begin to come in handy when the accuracy approaches 100%. Dragon doesn’t make typos; when it makes a mistake, it writes valid words, usually words that make sense  next to each other, but not the words you intended to say. We who have been typing for decades, will naturally look for typos when we proofread our text. It is all too easy for us to overlook that a wrong word has been used, such as “is” instead of “isn’t”. But chances are we catch it when we hear it out loud!

That’s all for this time, but I hope to be back with glowing praise when the accuracy approaches 100%. ^_^