This is a transcript of a podcast that discusses voice recognition and provides tips on making it work for you.
Speaker Key: PB: Phil Brown, DW: David Whelan
PB: Hi, it's Phil Brown and I'm here with David Whelan. Today we're going to talk about voice recognition.
DW: Voice recognition has been a little bit of a holy grail, I think, for lawyers, because it offers an opportunity for them to use one of their primary tools, their voice, to record information more quickly than they can type or write it. But voice recognition has had a mixed past and I think the best news is that it is getting better and better, and almost to the point of magic it seems when you use certain devices.
PB: It is not your grandpa's voice recognition.
DW: For sure.
PB: Back in the old days, you downloaded a very large package of software onto your desktop and spent hours training that software to recognize the various nuances of your voice, and there were certain words it was never able to recognize.
DW: It was particularly difficult, I think, because it meant that you were recording it in a particular way. I did a demonstration of it after having trained it my voice, but I was a little stressed out during the presentation, so my voice actually went up and all the training was no longer any good because it had been trained for a slightly lower toned voice. And so, it was very, very finicky. But now when you open up your phone or your desktop device, as long as you have the software on there, it's really remarkable how close the technology has come to matching most of the words that we use.
PB: Since you mentioned phones, let's talk a bit about Siri, which might be what a lot of people are familiar with. Is that true voice recognition?
DW: Oh, for sure, it's definitely voice recognition and Android has some add-on apps that do a similar sort of thing. But I think it's not the voice recognition that most lawyers would think of. Siri and other voice apps tend to be good at giving you directions or responding to a particular query because that query has gone into a very large database of other people who have asked the same information or asked for the same kind of detail.
PB: And, as we found out recently, people are listening to the recordings of your voice that are made when you talk to Siri.
DW: Right, that's kind of creepy.
PB: So there's no real confidentiality there and you wouldn't necessarily want to be using Siri for anything sensitive.
DW: We have talked about it on another podcast, but there was a recent case with a television company where the television was able to pick up your voice command so that it would change channels and do other things, but all of that information was also being sent off to a third party voice recognition service. So there are these large databases now being put together, obviously to help the people who are using voice recognition so that it becomes better for all of us, but it does mean that the things that you are saying into microphones may not only be recorded by you, but being stored elsewhere.
PB: Right, so there are a couple of big players and bigger software out there, and I'm sure a lot of people have heard of Dragon Dictates, and it's available for a number of different professions. They have special packages for doctors and lawyers and so on. It has come a long way since they started.
DW: It's incredible. Nuance is really a juggernaut when it comes to voice recognition and they have absorbed a lot of the smaller players, I think, along the way and it's a very, very strong tool. Even just right out of the box, you can start to have your words translated right onto your screen, but I think that the key item, if you can find it with voice recognition and, of course, Dragon Dictates has it, is this legal dictionary, because we have lots of terms of art that don't appear properly unless you've trained your package to do that. I found this with my daughter. Her name comes out as Chilean when you record it and so, even though it is a common name, it is not in the database, and so, when you think about Latin terms and other terms of art you might have in your practice area, having those built into a dictionary where it's already available will just mean that you're up and running faster when you get started and you won't have to trip and watch as you're transcribing your voice.
PB: And it can be a very useful tool for Smart phones. I have a free version of Dragon Dictates for my iPhone and you can dictate a large amount of text and then decide which platform you want to send it out on, whether it is an email, an SMS, a text, or any different way you want to send it - you just select after you've made the recording.
DW: It's remarkable how accurate it is. I've been really impressed. Even Windows has some accessibility options, which are really not intended for the general user, but the speech recognition in Windows is very, very strong. It takes a very small amount of training and you can be up and running, again, doing basic things. One of the key things to remember with voice recognition, whether it's on your desktop or on your phone, is that you always get a word and so if you're speaking into it, it never stops and asks, "Did you mean this word or that word?", instead it interprets it as "It sounded like you said X, so I'm going to put X in there." And so you really need to go back and read what you have transcribed or recorded because it may not include the words you expected. It's a little bit like your keyboard on your phone and sometimes we get words that we don't intend.
PB: No, for sure, there has to be, if this is going to be used to document, a lot of proof-reading and it's probably a good idea to have, if you're not a particularly good proof-reader, a second set of eyes go over it to make sure you're not sending something out that had a particular value in it and it goes out under a completely different number.
DW: Yes, that million dollar contract going out as a ten dollar one would be a problem.
PB: I'm sure, and so using it for things like a settlement and things like that, I mean, no matter how many times you've done it, I think you want to carefully read it a few times to make sure that things have gone well.
DW: What you'll find with some of the paid voice recognition packages too is that you can use a digital recorder, so, say you're driving up to see a client or off to court and you're recording on your digital recorder for a different matter or whatever, you can then upload that sound file later when you get back to your office and it will go through the recognition process from that digital file. You don't actually have to be sitting in front of the computer all the time or you can send that file to someone else who can then do the recognition for you, so it doesn't necessarily have to be something that you have to do yourself and to tie yourself to new technology. You can use the tools that you're comfortable with and then just have that digital file turned into the document that you're trying to create.
PB: Sure, and you can make notes using these various voice recognition applications, to make notes for yourself, and send yourself emails. You can get a lot of work done while you're driving, for instance, if you're dictating things to yourself that become emails later or documents later. You can certainly dictate a bunch of them in one file and then split them up and send them out as a number of different emails if you wanted to, so you're not messing with it while you're in your car.
DW: Keep in mind that if you're dictating on the road or outside your office, that your voice may carry and so it's even more so than with conversations. You may be getting the benefit of voice recognition, but if other people recognize your voice as well and what you're talking about, that could lead to some uncomfortable consequences.
PB: Sure, it's not the sort of thing you want to do in a public place and, I suppose, the other thing that tends to play into voice recognition is background noise.
DW: Yes, a coffee shop for example, not only is it not a good place to dictate loudly, but the clanking of plates and other things can definitely distract the technology.
PB: Sure, so have a look at voice recognition. There are a number of options out there and it might be a time-saver for your practice.
DW: Thanks, Phil.
PB: Thanks, David.