Interpreter 2.0 Mindset with Dr. Bernard Song

A speaker’s mic cuts out, the Wi‑Fi stutters, and you’re still expected to deliver clean, confident interpreting. That pressure is real, and it’s exactly why we invited Dr. Bernard Song to the show. He’s a conference interpreter with 25 years in the booth, a PhD in computer engineering, and the founder of GreenTerp Technologies, which means he can talk about interpreting craft and interpreting tech without guessing. 

We dig into the unglamorous problem that ruins performance fastest: audio quality. Dr. Song breaks down how automatic speech recognition (ASR) and computer-assisted interpreting (CAI) can act like a safety belt when the floor is messy, speakers talk at “uninterpretable” speed, or numbers and key terms get lost in distortion. We’re honest about the downside too: live captions can trigger instant cognitive overload and split attention. The takeaway is not “use tools everywhere,” but “train enough to control them,” including knowing when to glance at the transcript and when to ignore it so you keep tone, spirit, and meaning intact. 

We also zoom out to the bigger shift after COVID and the rise of AI translation in the private market. Some “just in case” assignments may move to AI, but high-stakes communication, trust, confidentiality, and human judgment remain non-negotiable. Dr. Song shares concrete, low-risk ways to practice at home, record your work, get feedback, and experiment with tools like TerpMate and GD Booth, including options designed for offline use to reduce privacy concerns. 

If you’re an interpreter trying to stay employable without losing what makes you human, listen through to the end, then subscribe, share this with a colleague, and leave a review with the one tool or skill you want to strengthen next.

EPISODE RESOURCES

AIIC Interpretation Checklist

SAFE AI: Guidance on AI and Interpreting Services

Transcript

Welcome And Why Learning Never Ends

SPEAKER_01 0:05

Hello Branded Bunch and welcome back to Brand the Interpreter. And if this happens to be your first time joining us, my name is Mireia Perez. I'm an interpreter, entrepreneur, and host of this podcast, where we explore the people, ideas, and experiences shaping the interpreting profession. Thank you for spending part of your day with me. Whether you're joining me on your commute, between assignments, during a walk, or maybe even while traveling to your next interpreter conference, I'm grateful you're here. Today's conversation is one that I think speaks to something every interpreter experiences at some point in their career. The realization that learning never really ends. My guest today is Dr. Bernard Song, conference interpreter, technology innovator, and founder of Green Chirp Technologies. Dr. Song has spent more than two decades working as an interpreter, but he also holds a PhD in computer engineering. That combination gives him a perspective on our profession that is both unique and incredibly relevant to the conversations many of us are having today. In this episode, Dr. Song shares how his love of language began at a young age, how that passion eventually led him into the world of interpreting, and why he believes that the moment we choose this profession, we also choose a lifetime of learning. Together we explore how interpreters adapt to change, how technology is reshaping parts of our profession, and why understanding new tools doesn't mean surrendering the skills that make us human interpreters. We discuss everything from challenging audio conditions and cognitive load to professional resilience, curiosity, and the importance of continuing to grow alongside the profession itself. What I appreciated most about this conversation is that it isn't really about technology. It's about mindset. It's about remaining open to learning, staying adaptable, and recognizing that growth has always been part of what makes great interpreters great. So, settle in and join me for this thoughtful conversation with Dr. Bernard Song.

From Math Kid To Language Lover

SPEAKER_02 2:46

Dr. Song, welcome to Brand the Interpreter. Thank you so very much for being here with us today.

SPEAKER_00 2:53

Thank you for inviting me.

SPEAKER_02 2:55

It is such a pleasure. I'm very much looking forward to our conversation and the information that you're going to be sharing with this audience.

SPEAKER_00 3:03

The pleasure is mine, as I said. I I can't wait to share my experience or whatever is useful for the audience of this podcast.

SPEAKER_02 3:12

Oh, it's going to be so much. And by the way, uh Dr. Song is joining us from where, Dr. Song today?

SPEAKER_00 3:18

Singapore.

SPEAKER_02 3:20

Singapore. And what time is it out there?

SPEAKER_00 3:22

Um it's 9 30, almost 9 30.

SPEAKER_02 3:25

PM, right?

SPEAKER_00 3:26

PM.

SPEAKER_02 3:27

Yes. So, so even more of an honor, Dr. Song, that you're joining us at bedtime. At least that would be my bedtime.

SPEAKER_00 3:37

My bedtime is actually uh two hours later. So I still have like a pretty pretty good time, you know, at that time.

SPEAKER_02 3:43

Pretty good time.

SPEAKER_00 3:44

Yeah.

SPEAKER_02 3:45

Well, thank you so very much for joining. And that's one of the things that we'll be talking about today is technology. And indeed, I am so grateful for technology for allowing me the opportunity to do this. Uh, being able to speak to interpreters from around the world, that is. So thank you again and welcome to Brand the Interpreter. Let's get started. My very first question, of course, as my audience knows, has to do with a fond childhood memory of yours. So I'd like to ask you, Dr. Song, are our careers often shaped by our childhood experiences, except we may not necessarily make the connection back then, right? Is there a memory now that you've worked in the profession for quite some time that connects you to your present-day work that you recall?

SPEAKER_00 4:30

Yes, of course. You know, um, many people are wondering, you know, because if you look at my titles, my education credentials is it says PhD in computer engineering, um, it says nothing about conference interpreting in there. Uh, but I'm a conference interpreter now. I I've been doing conference interpreting for 25 years and to today. Uh so it's a long career, um, but I think uh everything happens for a reason. When I was a young boy in school, um I'm always good at mathematics, you know, uh physics, uh chemistry. I love those subjects. Um my Chinese, I'm originally from Mayland, China, so Chinese is my mother town, my native language. I as a boy, I wasn't actually very good. I think that's a typical problem for most boys. You know, we were not so good at uh so-called the literature literature or arts subjects. Um, but then um my second language, English, is a subject in my school, and that gives me a chance to really excel because everybody starts from the same starting line. You know, um other other kids don't really have any advantage over me just because you know I started perhaps later than them. Um, so from there I started to deep develop the love for language, for English in the first place. And then when I learned English well, I started to love Mandarin Chinese because languages have common things. Um and so this this passion, this love developed uh itself along uh you know, until uh my middle school, my high school years, I got a chance to participate in a summer camp uh with students from the United States. And that opened my eyes because uh living in China, a country that speaks predominantly only Chinese, uh, you don't have the chance to be exposed to foreign cultures and any foreign languages. That was, I don't know, 30 years ago. So so um uh of course things are very different now. But back then, that was an eye-opener for me. I not only, you know, get to know more about the language, I get to know more about the culture, the Western culture. So going to college, I started to have my second major besides computer engineering and computer science. I started my

Moving To Singapore And Learning On The Job

SPEAKER_00 7:10

second major as an English major. And I'm so passionate about it. Um, I started to participate in a lot of contests. Unfortunately, I got the the national champion in the English public speaking competition as well as the English written contest in my college time. So that love laid a solid foundation, you know, um for this career. And not long after my graduation um with my bachelor's degree, I came to Singapore, which is now where I reside in for the past 20, uh, 23, 24 years. Um while I was studying for my PhD degree, I get a chance to uh participate in the social life of Singapore's society. And I got a chance to interpret, leveraging on something that I can I can do or I can do best between the two languages, because Singapore society is effectively bilingual or trilingual, you know, because uh 70% of the population here are uh overseas Chinese, they speak English as their mother town, a lot of them, but uh some of them can also speak Chinese. Although their race are Chinese, they look Chinese, uh their mastery of Chinese is not that high. So they still have to depend on their English for communication. And 20 years ago, when uh China started to open up, more and more people are coming out to the world, to the outside world. Singapore is one of the mice destinations and trade hubs and financial centers of the world. So the Chinese people are coming to Singapore, and there's a need for communication across the language. So I happened to fill in that gap, you know, without any training as a professional interpreter. I jumped in there and I started to interpret between the two languages. So most of my training are basically on the job. But that that's um that's attributed to my solid foundation back in China, in the English language, and with Chinese as my mother tongue. But I also realized that you know, to become an interpreter, bilingual is not enough. You not only have to be good at both languages, you have to know the ins and out to be good at converting them, and you have to also know the culture. So instead of understanding in-depth about Chinese culture, you started to know about Western culture and most importantly, Singaporean local culture, which is what we call a raw jack, a mixture of Eastern culture and Western culture, all right, because of his uh colonial you know status. So I'm I'm very fortunate, you know. I was at that time to fill in that gap. Uh while pursuing my PhD degree as a computer scientist, I started to boot my career as an interpreter, conference interpreter, legal interpreter, etc. I I did some uh very few um but but some community interpret interpreter jobs. So so that's how I developed a career. And then before COVID, I was so busy being a conference interpreter, I quit my job as a computer scientist because I decided I can't really focus on both of them. The subjects are really wide apart, although I really benefited as a conference interpreter from the training that I receive as a PhD in computer engineering. The, you know, the the qualities about learning and learning fast, those are very important. Uh, because this career as a conference interpreter is similar to a computer scientist. The moment you choose this, you gotta keep learning. And there is no stop, lifelong learning.

SPEAKER_01 11:19

Exactly.

SPEAKER_00 11:20

Because the world develops so fast, and as an interpreter, you got to learn new knowledge, and every day your challenge is different. You're interpreting for chemistry conference, medical conference, legal conference, and the knowledge in each sector keeps developing fast. Now with AI, you know, we all know how the world is changing uh so fast. But COVID um is another blessing in disguise.

SPEAKER_02 11:45

Yeah.

SPEAKER_00 11:46

COVID had helped me connect the dots behind me. A computer scientist and a conference interpreter. So when our career is going through this pivotal moment of transformation, we need to get away from face-to-face meetings. We need to continue to work online in a digital format. How can we be relevant? So that is where I bring the technology in to help interpreters. And so today I'm working on you know creating uh CAI2s or using technology to help us as interpreters.

SPEAKER_02 12:25

What were you seeing that were obvious pain points for interpreters

Lifelong Learning As A Core Skill

SPEAKER_02 12:30

to be able to provide their services adequately?

SPEAKER_00 12:35

Yeah, that's a very good question. Because before COVID, or like before 2018 uh time, um we were like, as interpreters, uh, very few of us use technologies to assist us. It's the traditional techniques that we were trained in schools or acquired uh on the job, uh, taking notes with your pen and paper. Everything is relying on your listening capability and your brain, your processing capability. So all the trainers, all the seniors were telling us listening is everything. If you couldn't get the good input, you can't you couldn't process it, understand it, you can deliver rubbish, right? So so the input is very important. Um, but around 2016 and 17 time, I started to think uh maybe the technology is there. You know, I've we started to see autonomous automatic uh speech recognition technology came into being, and the accuracies are increasing, and it's almost a live streaming type of ASR. And that really helps because it shows right in front of us um what the speaker is talking about. So besides what you're hearing, you also get some kind of visual uh reflections right in front of your eyes, and your hearing, you know, sometimes you may have illusions, sometimes your hearing is still better than ASR. But as ASR can be a good complementary thing, um you know, it's just like a safety belt. When you fall, it really saves you from falling. Um and so I started to experiment using ASR back in the 2016, 2017 time, when the first generation ASR technology came out. And we started to, not just myself, my colleagues, the the teammates that I work with in the booth, we started to test that and we started to try to use it. Uh, it's a long process until today you see some products available in the markets, but then it's it's it has gone through a very long process, and I firmly believe that's the trend because along the way we started to use it, we started to train ourselves. It's not something that is miraculous that you you um you're starting to use it and suddenly you you become very efficient. Not really, you know. Everything in our career, it's it's like a struggling, it's not really natural, right? When we do simultaneous interpreting, you've got to constantly fight um within yourself. When you're listening, you're not talking, right? And then when you're talking, you're not listening, but you need to coordinate these two, and you need to do these two together. Now I give you another modality of input, which is a moving text right in front of your eyes, which may or may not be the same as what you hear. So you gotta coordinate between the two modality or even three modalities of input. Um, some people may say it's a cognitive overload at the beginning, but I believe with um proper training over the time, you become proficient, just like simultaneous interpreting can be trained. Our brain can be trained to do two things at the same time. Now we our brain can be trained to do three things or even four things at the same time. Hotter training.

SPEAKER_02 16:25

You know, okay, so I I have a couple of questions as follow-up, but I do want to go back to what you what you were experiencing. It sounds like it mainly had to do with the with the audio component, which I think many of us in across all different sectors of interpreting, we can we can relate to, relate to the fact that I can interpret what I cannot hear. And at a at a very basic level, uh, if I want to call it basic, but a different level, perhaps when it came to individuals opening up their laptops and you know, simply turning on a computer and starting to talk and relying on that built-in microphone and speakers. And it was it was horrendous. It was it was an awful experience, which is why I think for many of us in the community setting, our our initial thought was there's no way this is here to stay. It's awful. There no one knows how to work with the technology at its most basic level, which is you know, proper uh input output of sound through a device. Um, it was really difficult for us to get individuals to wear headsets and use, you know, the microphones and try to explain the reasons why. So it sounds very much like the the main drive for your passion,

COVID Connects Interpreting With Tech

SPEAKER_02 17:40

your interest in this topic had to do with maybe you experiencing that piece yourself, that it was it was difficult for you at that level at a conference interpreting with individuals that you know, I don't know, perhaps had the funding to fund more technology. You were still experiencing audio issues. Am I correct to say that?

SPEAKER_00 18:02

Yes, absolutely. I mean, I very much agree to what you said. Umbe I have to make a disclaimer here. So um for staff interpreters working in institutions like European Union or United Nations, their situation is very different from the private market interpreters. Uh, I work predominantly in the private market. Uh, I do work as a NAIC member in uh United Nations conferences or international organizations, uh, you know, as a what they call a freelancer or uh freelance agent. There are many names to that. Um, but I'm I'm not really working full-time in there. So the working conditions are quite different, and uh people have different uh aims. Um for institution interpreters, of course, they wanted to maintain a very good working condition, and a lot of times they can by collective negotiations, but for the private market, um the the balance between the buyer and the seller are not really always there. And although very much we wanted to have a good working condition, the booth, the proper equipment, um we don't have them. And and for community interpreters, I think we're all aware, you know, uh, because of the budget constraints, a lot of times uh interpreters are forced to work in situations where they can't hear very clearly. You know, the audio quality coming into their ears are really challenging. Um and especially after COVID, right? When COVID came, um, we were using video conferences and people are using their onboard microphones from the computer. They're not having a proper microphone, they're not having even a headset microphone for them. Uh, these are all challenges. And I happen to find um that there are ways to use technologies, at least to mitigate the problems. These are not applicable to institutional interpreters, I have to declare. Um, but this is to help private market interpreters where sometimes we don't have a choice. The job has to go on and we gotta deliver a good job, right? And at the same time, you gotta we gotta protect our hearing. So, how do we make ourselves uh a good sound? Or are we able to leave with whatever that input is? One thing I realize is that during the COVID time, when people are dialing in from you know uh choppy Wi-Fi's using um bad microphones for human hearing, it's very challenging. But if you have the ASR, it can basically get more juices out of that bad sound. So you will see uh what was there.

SPEAKER_02 21:05

Dr. Sung, I don't mean to interrupt, but could you break down ASR for those that are unfamiliar with with that acronym? Sure, sure.

SPEAKER_00 21:13

By ASR I mean automatic speech recognition. So it's like the Zoom, uh they have a butt-in um transcript or closed captions. Um you can have that in Microsoft Teams meeting. And even if you're using uh MacBook, uh Mac OS or Windows, nowadays they have the built-in uh closed captions for certain languages, not for all the languages, but the major languages they have that already. Um, and to a certain extent, it really helps. The machines also struggle, but you know, in a situation where human interpreters cannot tell what the speaker is talking about, it's like really robotic voice. I find that the ASR transcript. Still makes sense. I know there are arrows. Even if like you're in a perfect audio condition, there will be arrows. But then the keywords started to come out. And as experienced interpreters, you're able to piece them together and tell a story. Because at that time, keep the communication going on is more important. And getting across that meeting is more important during that time. And after the meeting, other booth colleagues are asking me, like, how do you do it, Burnett? We can't work. How can you keep a steady output? We we can't tell. But we we we tune into the Mandarin channel and we hear you giving stable output. So that makes a difference. I'm not saying you should use CAI tools or ASR2s all the time on every job. But in

Audio Pain Points And ASR As Backup

SPEAKER_00 23:00

certain occasions, on certain occasions like this, it really helps. It's like a safety belt, like I said, you know, you would need it. But in order to use it, what you need it, you need to be prepared, you need to be ready. And to reach that readiness, you need a long time of training. So when you need to use it, you can use it freely rather than you know making that a barrier, making it a distraction or a cognitive overload for you.

SPEAKER_02 23:31

Yeah, that was the first thing that was coming to mind. Yes, absolutely. I was thinking, I even wrote down like that, it sounds like there would be cognitive overload, right? Um, it sounds like there the the issues, there would absolutely be the split attention issue for an interpreter. And um, I really like that you called, you know, it's a it's an additional modality of input that that the interpreter would be experiencing. And I think that when I've experienced it in in observation for someone that is untrained, I can see how it becomes a barrier. But very similarly to I think someone in in one of your um articles uh shared this in a comment, which is exactly what I was thinking. It it was very similar experience to note-taking when you're first trying to learn the note-taking technique, listening and writing at the same time, right? And so, but but in this case now we're talking about a third uh potential if you're doing a simultaneous. So those are some of the things that were initially coming up for me as you were talking about uh Dr. Sung called it the uh Kai C A I, but uh to break it down for our listeners, it's the computer, a computer-assisted interpreting. Is that correct, Dr. Sung?

SPEAKER_00 24:54

Yes, yes.

SPEAKER_02 24:55

You saw a shift, we all did during COVID and the use of technology when now it sounds like it was oh, so long ago. Right? Like for us, it just seemed like, oh, we went through it. It's like a tidal wave, it hit us, and then we, you know, we did our thing, and then here we are still on top. Uh, but now we're getting hit by newer technology. Walk us through what you're seeing now in terms of potential similarities or differences when it comes to the shock of the interpreting profession and adapting to the new technologies. Are we moving at a faster, more positive speed because of COVID? Or are you seeing in your own experience that we're experiencing the same emotional roller coaster as we did during COVID times?

SPEAKER_00 25:44

I I think I agree with you that um after COVID, you know, uh, we have gone through that that uh intensive period. Um the next challenge or the next impact is AI. I think everybody knows that it's impacting not just us, every every job, every career. Um, we all heard about the stories of uh big layoffs from the um the MNCs or those IT big giant IT players like Oracle, right? They laid off so many people uh suddenly. Um, and a lot of them were programmers who basically developed the codes or developed the AI components, and they got laid off. Um so comparing with them, you know, um, I think that's something we have to see coming. Um, but how do we prepare ourselves? How do we deal with it? As human, that's the resilience that separates us apart from animals or from machines or AI, whatever that is. So we got to demonstrate our resiliency, find the uniqueness of being human, and uh take the challenge, overcome it, right? If you if you resist AI, it's no good. It's like what happened when cars are taking over horse carriages, it's even like emails are taking over our mobile phones, right? So people will move, the mass um population will move with the trend. Uh, the capital will push for it, and and we already see this happening. You know, a lot of jobs were previously done by human interpreters, are now moving to AI speech translation programs. Um, that also gives us a chance to review what is really our position as a human interpreter. What are we different from uh the AI? In the past, a lot of the interpreter jobs or interpreting jobs can't be uh done by AI because AI was not an option. But frankly speaking, we know a lot of jobs were not that important. It was just there, you know, as a form factor as to show that we provide accessibility. So called this accessibility, language accessibility there. Uh almost everyone in the room can speak, for example, good English, or they can communicate in the same language. Um, but with interpreters there, it really helps. Or maybe for a conference, only one guy or two guys picking up the headsets, right? We've seen too many of such cases. And now maybe these jobs will be gone because people will see AI more efficient in terms of fitting that gap. Um, but humans will still remain relevant in the high-stake scenario when you have reputational risks and other risks. And of course, the human touch is very important in there. But it doesn't mean that we're fighting with AI, okay, with whatever we have. We, yes, we will perhaps compete with AI, but we also want to make sure that we compete with AI based on empowerment of AI or other technologies we have at our hands, right? For example, when some speakers in private market, they just speak at lightning speed. We call it uninterpretable speed of speech. Now, in the past, what can you do about it? Focus based on your hearing, filtering out things, and then try to you know deliver as much as you can. Um, what if the information is very dense? The guy's not bullshitting, right? He's actually saying things that are dense with valuable information. You can't really, you know, do the trade-off. So again, coming back to ASR tools or Kai tools, it can really help. It will give you everything, it's a word-for-word um transcription out. And sometimes you may be a little bit behind, but you get everything in front of you, right? So that really helps, and you will find interpreters with the proper assisting of uh of ESR2s in that kind of situation with you know breakneck speed of speech, they can still handle it pretty well. This is unthinkable before when you're using such a tool.

SPEAKER_02 30:45

In your article, Dr. Song, you talk about refining uh the Kai skills through accessibility uh to training and tools. Yes. Conference interpreters, as a blanket statement, um, I recognize that there are not not all, but but when we think about conference interpreters, we think about conference interpreters often having the access to these resources, to formal programs, to colleague networks, whereas to oftentimes community interpreters, um, those of us that are working out in the community services, don't. So what would you what would you say to it to a community interpreter, for example, that that is thinking about these tools, that is seeing maybe someone that is coming from, let's say, the conference world into a community interpreting scenario that is utilizing these tools and they've no idea where to even start.

SPEAKER_00 31:46

Yeah, I think for community interpreters, it's even more helpful to use technology uh than conference interpreters, because conference interpreters usually work in a pair of two, even in consecutive interpreting occasions, they are able to command uh, you know, a team of two at least working together from the client. But from the community interpreters perspective, most of the time people are working alone, right? Even if when you are doing a one-hour

Cognitive Load And Training With Tools

SPEAKER_00 32:20

uh parent meeting with the teacher, uh, even if you're doing simultaneous interpreting for that kind of occasion, you don't have a booth for you. You're perhaps using um tour guide systems, um, but naked year, you don't have a good audio input. Um and and nobody is there to assist you as a partner if if you you miss certain information, miss certain word. So, really, technology can fill in that gap. You know, in the first place, how can you use technology to provide a better live feed? We're able to do that now, you know, in the using tools like a broader kai2 is not the traditional narrow sense kai to that is uh automatic speech recognition. Um, we can also have tools that can bring the live feed to either hardware or software to your years. For example, a lot of the conference rooms or even school classrooms are equipped with Zoom meetings thanks to COVID. You know, they already built that in. And so what's helpful is that as a routine, they just you know, turn on a computer, dial in the Zoom meeting, and uh interpreters can be sitting either inside the same room or the next classroom, next door classroom, uh, but listening through Zoom. Now you'll get better sound quality input. You gotta see what's going on. It's much better than you sit in the room naked year using a tour guide system. So these are these are simple tools that can improve your working condition. And secondly, when it comes to working alone without a booth mate or without a partner, the technology can really help because it really serves as something, you know, like a prompt for you. And the ladies' tools not only give you the ASR, the speech recognition result, it also gives you the reference translation result. Right? I know it's far from good human translation, but what you can do is you can pick up the word. Okay. For example, if you're in working in a scenario like a like a hospital or or medical settings, the doctor is saying something that you kind of like blurred out your brain, you can't really get it. But then most likely in the full text translation part, you will see that keyword being translated. And that is like your booth partner writing out that word and they know what you wanted, right? Uh, and the numbers are coming out. So these are really helping community interpreters when they have nobody else to depend on.

SPEAKER_02 35:17

I really like that actually. Uh, and it's exactly that example that you just gave, which is it's like it's like having that second person there writing out that translation. Uh, obviously, in my mind, I'm thinking this isn't something that you one as an interpreter should just turn on and and start trying to utilize it during a live interpretation if you've had no, you know, no, no training or no experience with it. So for those of you that are listening that have never had that experience, don't go into your next session and turning them on and then attempting what Dr. Song is explaining, like, oh, great, I'm gonna have a partner to help me because that might that might get messy.

SPEAKER_00 36:03

Exactly. Right, yeah, that will add to your cognitive overload.

SPEAKER_02 36:06

Yeah, don't go starting that experience. Although we do want you to start in smaller steps and being able to access these tools, obviously to one's advantage. Um, Dr. Song, you you've mentioned too that free access to trainings removes the paywall, but it doesn't remove the learning curve. So from from someone that is starting from zero, which is uh, in my opinion, perhaps many of the community interpreters for a variety of reasons, what is that absolute smallest yet safest way to experiment with with these tools? What should they be doing or could they be doing in a way in which is similar to a live session, let's just say?

SPEAKER_00 37:00

Mm-hmm. Um, so this is also part of the changes of uh continuous uh professional uh development, CPD. Uh in the past, we gotta attend courses and trainings to improve. Um, but don't forget that you, as an interpreter, you have your daily exercise and your daily training. And in order to make that effective, you can actually uh record what you have done in a simulation, for example, and then replay it to yourself. And you will find a lot of times um you can't bear with what you interpret. It's like, oh my God, I've made such so many mistakes. You know, I don't like hearing my own voice. But that's a good thing. Uh start to love yourself, start to embrace your mistakes, um, and and improve in there. You know, if you don't like your voice, next time remind yourself uh that you can improve. Um, and some of the skitus um do provide the recording functions now. Now that's uh that's not a simple recording where you can just turn on your phone to

AI Disruption And Where Humans Win

SPEAKER_00 38:10

record, because we know when you're recording simultaneous interpreting, you need to record the floor, and you also need to record your own interpretation at the same time in a synchronized manner so that later you can play it into two years. Left is the floor, and right is your own render rendition, and then you compare. This would be called an effective uh you know exercise uh or training. And and so you can leverage on such tools to do that. Baby steps, play a YouTube video, you select the YouTube video, or I think there are abundant uh training resources like uh uh audio recordings or video recordings from the European Union, um uh European Commission. Uh, they have that in the database. It's open to all, I believe. Uh you can download and you can start to train yourself with those uh you know recordings, or if you are a community interpreting uh interpreter, you can um record whatever you you have done um in the job and then play it back to you because it's offline on your computer. There is no privacy issue or security issue. You take it back for your own review purpose. Okay. Um and then for online training, you can find a partner. Because now there are tools that are available for interpreters to use virtual booths. Um, and then you can monitor each other while you're interpreting, for example, for a live broadcast, a live podcast, whatever that is. You're listening to the whole same thing, and you're interpreting your booth partner is monitoring you, and then you give feedbacks afterwards. So these are simple ways to improve our you know our our our uh skills. Um uh, but at the same time, you can also learn to work with the tools, the technologies. I don't have you know a so-called systematic way to teach people how to use the tools or to train you how to use tools. It's just, you know, it basically feels like at a beginning a site interpretation job. Yes. Because if you're giving a site interpretation job, right, five minutes before the, or maybe one minute before the meeting starts, they throw you uh a paper uh speech uh transcript. And then you don't have time even to go through glance through. So you've got to follow sentence by sentence, right? Uh, but you you don't want to go into every word, right? That's the difference between human and AI. You don't go word for word, you go for the whole sentence and try to get the meaning out. Now it's even more challenging because imagine now this this whole sentence is coming out. You don't see the later part, it's coming out right in front in front of your eyes and it's jumping, and sometimes they may even change itself, and there may be arrows too. So so these, of course, will add to your cognitive load. So you you start to you start to uh depend on your listening, but at the same time, you're also trying to extract information from there. If you feel it is too much uh of uh of a cognitive load or too challenging, you may move your eyes away from it. But every now and then during your training, it's good for you to look at look at it every now and then. And and the key thing is you have to know when to use it and when not to use it. If it presents too many arrows, too distractive, you just stop for the moment. And next time, it's a training, so you gotta restart and try to see how you can better use it. It's an experimental process. Everybody may have different ways of using that tool uh to your cognitive advantage. Um, for me personally, I've been going through this journey of seven, eight years already, since the first time we started to use some simple ASR tools now to a fully developed Kai tool. It's a long road. And I I have to say that sometimes I have to move my my eyes away from it. I don't want to be dragged in there. So I know when I can move my eyes and I I listen. All right. And when I need to seek help, I I would uh seek for help. So it's a two. Uh, you know, two to a human is like when you need it, you need it. The two will make you more powerful, but you don't get dragged down by that two.

SPEAKER_02 43:00

No, I love that. I think that's exactly how in in any of the new technologies that are being rolled out that that uh may or are having impacts in our profession, that's how they should be looked at. As you're having this conversation, as you're sharing this uh information with us, I'm I'm envisioning the very first time that I turned on um closed captioning on a virtual platform um session and it it was an immediate uh cognitive overload for me. My brain wanted to transition to um a site translation. And then, and then when there was a word in there that I recognized that didn't fit, it just it didn't fit with the context, I I realized I had stopped listening. And now I didn't know what the actual word was. And so it was an immediate, you know, like, oh, let me turn this off. This is not, this is not good for me, right? And so I was envisioning that specific moment for me that I thought. Thought, you know, oh, this is how can anyone say that this supports you, right? This is way back a few years back when it was first, you know, being rolled out for accessibility on the platforms. Uh, but I'm I'm also thinking, you know, obviously, like any tool, we've got the option to experience something like that and then bury our heads right um in the sand and say, this is this is not helpful, or learn how it can be helpful so that when I do have the opportunity and can use it, I it's there, it's accessible, and I know how it's going to support me because I've recognized, you know, this this is a tool, like you just said, and I know when I'm going to use it and when I'm not. I also think, Dr. Song, and I don't know if this is something that maybe in your research and your conversations is come up with other interpreters, but many of us recognize that, especially in community settings, not only, but in community settings, we often talk about ensuring that we are not just interpreting meaning, but also remaining faithful to the spirit and the tone of the message from the individual that is speaking. Have you come across issues with uh those specific elements in using uh computer-assisted interpreting, in which an interpreter maybe loses the spirit of the message because you know it's it's it's more uh systematic. I don't know if I'm using the right terminology there, as opposed to really attempting to make a connection with the speaker's intention and the spirit of what they're trying to share.

SPEAKER_00 45:51

Yeah, um I find that all the time, but again, a two is a two. It depends on how you use it. Um well, in my experience working as uh not as a community interpreter, but uh interpreter for motivational speakers. I do a lot of motivations uh speaker series, uh uh trainings, uh interpreting. And uh one of the important factors is you mirror uh the tones, um, the intonations, you know, uh the passions of the speaker. So it's like what you say to bring the spirit of the speaker out because what they see on the stage is somebody jumping, you know, shouting very high, and then they hear a Dow voice in their ears, it's a mismatch, you know. And so we gotta ensure that we deliver the same tone, the same feeling to the participants in a different language. We're like the voice of the speakers. Um, so in that kind of scenario, you tend to you know stay away from uh the transcript more and focus more on you know the the the feelings, the the spirits. Um and every now and then if you feel that you can't catch certain words, you may glance uh at the two to seek for some help. Um but again, you know, like I said, if you need it, you need it, but you don't get buried in the words, in the sentences. Uh because at the end of the day, that's where we are different from AI interpreting. AI interpreting is still word, we call it verbatim translation. It's word for word, word by word, but human interpreting is not, because we all know um the target language is longer than the original language. And if you want to fit your target language within the same period of time or even shorter time, you gotta be succinct, right? You gotta choose the right word. You don't have the luxury of like explaining a long sentence in an even longer target language sentence. So that's the trade-off we are constantly fighting with as interpreters as well. And then talking about the you know, the the previous point that you mentioned when you when you first get to use a Kai too, you get stuck with certain words, it's not right, or you just can't figure out what that the meaning of the word is. It actually happened without that tool that we constantly get stuck with in certain words, and we we wanted to get that figured out and then deliver a beautiful, perfect sentence, but we can't. I mean, people say it simultaneous interpreting is is an art, or even consecutive interpreting is an art of regret. We we just have to give up, you know, um, and move on because the sentence

Why Community Interpreters Benefit Most

SPEAKER_00 49:10

doesn't stop. The speaker is going on, you gotta continue and then do better for the next part. Um, and and similarly, when the when you're struck by that too with certain words that couldn't fit, you just have to tell yourself, get rid of it. All right, for the moment. That's a training, you know, that's the spirit. You move on, and you move on. And then hopefully, you know, later when your partner is interpreting, you got a time to rethink about it or check it out. What is it?

SPEAKER_02 49:41

Yeah, yeah, because that's how we refine our skills. It's so true, right? Like we like I mentioned, we don't. I think what I'm really appreciating about this conversation is recognizing, and for our listeners, hopefully, too, that that technology should not be an all or nothing thing. Meaning, you know, we shouldn't, we shouldn't think about it using it all of the time. We must use for every occasion, we should be using, you know, technology. Whereas what we should be looking at it exactly as you're you've been explaining to us this whole time, is looking at it as a tool. Except if I'm going to pull out that tool to help me, I have to know how to use it. And so in comes that that level of training that is going to allow me the opportunity to at least at whatever level be able to know how I can use this tool to my advantage so that it can make me better, right? Not the other way around. So I think that for for the very first thing for our listeners is that mentality shift. We hear about technology and AI and all sorts of other uh key words out there, buzzwords in our profession, that sometimes it's like, oh, I'm so sick and tired of hearing all this conversation about AI. Uh, you know, we want to we want to go back to speaking about all of the other topics uh that relate to our profession. But technology is here to stay. And so we we have to get comfortable with these conversations. We have to get um familiar with the terminology that that is now new terminology in our profession that we didn't hear maybe before. For some of us, it's the first time we're hearing high, right? Computer assisted interpreting. And so it's super important for us to not only know about it, but also delve into how we learn more about them. Dr. Song, one thing is us as individuals learning about um these tools and utilizing them. But earlier you spoke about the interpreter that is the employee, right? The interpreters that are working um within the compounds of an organization. And sometimes it is those individuals in the organizations that are having conversations about bringing in these tools. How can an interpreter be able to ask the right questions about an AI or a CAI tool as it relates to their profession? What questions should they be asking technology vendors that are coming in to be able to support not just their own understanding, but the understanding of colleagues that are making decisions about break bringing in these tools to organizations? Is there any suggestions there, any feedback that you can provide an interpreter that is trying to mitigate these topics within their organizations?

SPEAKER_00 52:53

Um, I think for interpreters who work as in-house uh staff interpreters within institutions, um I do see see some major pushbacks uh of using technology tools, uh mainly because they would be thinking, you know, of these tools. Number one could be distractions, cognitive overload. It will make our work harder. And then uh if it really, as you said, you know, after training and working with it, after many, many hours, we get to use that and things become easier, we come, we become more empowered. What if our employer would you know ask us to do more because the job becomes simpler? Um, you know, and then uh it's just like what happened to the the programmer's world, right? When AI becomes so powerful, uh what a team of 10 people can do in the past now can be done by one human programmer plus 10 AI agents, so they fired all the night. Yeah, that's that's not good, you know. So so there are really critical questions to ask to um if if institutions are trying to bring these tools, what is the real purpose behind? You know, are they really helping the interpreters, making their work more efficient, or are they just simply looking for um replacement tools or you know, efficiency improvement while cutting down manpower? The answers will be very different, depending on the answer um from the institutions. So I don't really have an answer to to this question. It's up to the colleagues who work in the institutions to find out. And I'm totally with them, you know, their their concerns are very valid. Um, because if the institutions are trying to lay off interpreters um by saying, you know, we have to's that are efficient. Uh, we don't have AI to replace humans, but we have to's to make humans more efficient. Maybe we'll need less humans and more twos. Um, that's not something we wanted to see. Um, so that's why I made a disclaimer at the beginning of this conversation that whatever I introduced is for private market interpreters where we don't have a choice, right? Uh we if if the client decided, the client has already decided if they want to use AI or they want to use human. All right, they already make a decision and they want both. I can tell you, most of the clients are sophisticated. Okay, don't expect your client nowadays to say we're gonna stick up to human interpreters and we will never use AI. It's not true, it's just like human consumers. We like Nike, we like uh LV, Louis Vuitton as well. So in different occasions, we use different brands, we use different apparels to dress ourselves. So for occasions where human interpreters are needed, the client will hire human interpreters. But for those occasions, according to their assessment, AI is sufficient, all right, or then when they don't have any budget for human interpreters, AI

Safe First Steps To Practice

SPEAKER_00 56:20

would fill in that gap. It's actually an expansion of the market for the private market. So it's not a bad thing, it's actually a good thing. You will see a broader market, you will see um in this process, more opportunities will present itself. I like that. People started to try out AI and they realize AI cannot do the job. We still need human at the end of the day. So that brings the job opportunity back to human.

SPEAKER_02 56:48

Absolutely. And you've mentioned it a few times, and and I recognize that for many of us too, we recognize that those high-stakes um meetings, those meetings that we recognize as the need to be private and confidential, you know, there is very specific moments in which the interpreter, human interpreter, must be there. And I think as as we continue to talk about these topics with the general public, particularly those that are making decisions internally in organizations, it's good for us to use those keywords that are going to help the general public have a better understanding as what does high stakes mean in your organization. Safe AI has packaged together ways in which we can have these conversations with vendors that are coming into uh the public sector and trying to sell their product. Uh, what are those, what are those questions we should be asking that are relevant to the work that we do? And so um, I'm gonna insert that information in our episode notes so that if individuals are interested in just reading a little bit more about what are those questions that you should be asking, you know, you have the opportunity to with a with a document that can guide you, but I do think it's important that we understand how we can begin to differentiate the work that we do with something that machine can handle. For example, you just mentioned, you know, the the the type of meeting, maybe or event in which just in case, right? Can I have an interpreter just in case because there is enough of the language there? We get that in the community setting as well, in which the interpreter is just sitting there just in case, because the parent wants to, or you know, the the client wants to use the English language. They want to be able to communicate directly, but has someone there in the background just in case. Maybe those are the moments I hear a lot of interpreters saying, oh, we just sit there. And oftentimes they just do fine and we don't have to work there. And it's, you know, they the interpreter wants to work. Well, maybe that's you know, where we could use technology for the just in case moments, right? It I think it does really depend on that perspective.

SPEAKER_00 59:11

Yeah, I I can give you a reference. I think IEK has uh an AI question uh checklist, um, which is a flow chart um with questions. Basically, uh it's launched by AIK Science Hub AI work stream. Uh you can find it on LinkedIn. It's already publicly available. Uh, so that workflow is basically what clients need to ask themselves, what questions they need to ask. And then following that flow chart, they're able to identify whether the occasion, uh, you know, for that occasion, human interpreters are needed or AI will be sufficient. So it's really helpful for you know clients who have very um who don't know whether they need AI, you know, they just make a decision based on their hunch fittings. Now they get something more scientific, all right? They can actually follow the checklist to make an informed decision.

SPEAKER_02 1:00:10

Thank you. I'll make sure that we can include those in the episode notes for our listeners. Dr. Song, this has been incredibly practical. And before we get to the close of today's episode, I want to bring it back to you personally and leave our listeners with something that they could hold on to. Um, if you wouldn't mind giving us three small concrete actions that our listeners can take in the next two weeks, let's say, to be able to start evolving with technology on their own terms, as opposed to being felt like they're being pressured to.

SPEAKER_00 1:00:45

Yeah, I think um everything has to have a start. All right. After we talked for an hour, um, I hope more interpreters will give it a try, give it a spin. Um for my company, GreenTurp, we did something to bring down the barrier of usage because we found that if you uh have a paywall in there, people would say, people will give them the excuse, oh, I have to pay, you know, to try it. I will probably not try it now.

SPEAKER_02 1:01:16

I've done that.

SPEAKER_00 1:01:17

Yeah, now I've removed that barrier, so you don't have an excuse to try it, right? So uh the to not to try it, um go and download um whatever two that you want. You know, we have the turp mate, which is a ski two. Um, it's it has the narrow Kai C AI sense in there, but it also have a broader um two set that will help interpreters to get you know the live feed onside there, uh more functions or or recording, you know, the the two-tracks recording, like I mentioned. Um, we also have the GD booth, which is a uh virtual booth tool if you want to practice and monitor and give feedback to your booth partners um of your of your interpretation, you can use that too by comfortably sitting at home. You know, you don't have to sit together next to each other, make it so formal, you know. So you can just arrange a time like nine o'clock in the morning, and then let's do a one-hour uh practice so the two of you can actually use that too for free to uh to do uh an exercise and give feedbacks or reviews. I I think you will find that really helpful because that's a that's very uh that's something that I find very useful. Um, even by listening to myself or having someone listening to me and giving me feedbacks will help me improve really quickly. Because I've many uh blind points, and I believe every interpreter has blind points. Absolutely. And you just hate to listen to your own voice. So you whatever you have recordings, even if you do the recording, you don't want to listen to it. Now, if you have someone there live listen to you and give you feedback, uh at the beginning it's uncomfortable, but then you get to know

Institutional Risks And Vendor Questions

SPEAKER_00 1:03:10

your blind points, you gotta improve in there, and trust me, you will improve very quickly. So, so these are these are my tips. If um if you encounter any difficulties, uh please reach out to me. You know, I I would also I would be very happy to help anyone as an interpreter. Um, I'm the in the AIK uh science hub technology work stream, I'm the convener of the work stream. So constantly I receive questions from Aik members. Uh, if they have any technical questions, they can send it to me. But as non-Aik members, if you have any technical questions, you can feel free to contact me and let me know. You know, whatever I can, I will try to help you address your questions. I hope this will bring down the technology barriers as much as possible so interpreters get to start. And then remember, don't give up. All right, you will have that frustration, that difficulty. You know, just like anyone who first tries simultaneous interpreting, you will find so hard, right? How can you listen and then speak at the same time? Can you survive the first five minutes? Right? So this is this is a long journey, but um trading something useful. Okay, when you have the two to assist you, um it's like you plus AI versus AI. You will never be beaten by AI. Right? So it's what it's worth investing the time and efforts to do something like that. And then this is the big trend. Okay. I I do advise everybody to read more, to understand the big trend, to understand more about the market dynamics. We can't be like before, just close our eyes and then just do our jobs as interpreters. You gotta know what the market is doing, where is the future direction? Um, because the market is moving very fast, right? For the past 100 years, uh is the traditional CI consecutive interpreting or simultaneous interpreting. That's it, no new tools, right? It's the console's upgrading itself, right? But then from COVID until now, you see oh, so many generations of new technologies. And even for consecutive interpreting, now we have the tablet compute uh interpreting tools where you are taking your notes, but you also get those ASR results and full translation results. You have a virtual booth mate with you there, and so that you know you have something to depend on.

SPEAKER_02 1:05:43

Absolutely. Invest in yourself. I mean, that's really what I'm I'm hearing at this point. It's like we there's no going back, right? Like we either become the uh interpreter 2.0, or we're going to get left behind. And there's so much. Much more that is out there to help us improve and help us get to a place in which we feel comfortable using the tool and have expanded not only our knowledge, but our performance or technique to make us even more employable, as far as I'm concerned, more uh more hireable.

SPEAKER_00 1:06:20

Um, this is also very precious because a lot of concerns from senior experienced interpreters is that they worry their skills will degrade after they get to you know rely on tools. Not necessarily so. Again, it's it's there they there is a curve in there. So you follow that learning curve. When you first learn something is really difficult, right? And then when you are really good at doing it, you become dependent on it, you become very easy. But then that will also have the draw because you have to you have to separate yourself from the twos. So that even when you are faced with challenging situations, a hard job, without twos, you can still survive very well. But we use twos to make our life easy. Don't forget about that. All right. You it's if if there's a two to make your life easy, you want to use that too to make your life easy. And and without that too, you can still live your life. That's the whole purpose. You it's not like I want to remain in the hard mode all the time, right? I wanted to ride a horse carriage every day instead of taking an airplane or drive a car. No, that doesn't make sense, right? I wanted to use that too to make my life easy. Yes, when you're good at it, it will make your life easy. You will do uh after a whole hard day of work previously, you're exhausted. Now you don't feel exhausted. You still have the energy and the brain power to go back home to spend time with your kids and your spouse. Have a good life. All right. And keep your skills. You also have to occasionally keep your skills. So the next day, if you you don't have any of these tools, you can still perform good. So that's the balance we're trying to, a dedicated balance we're trying to achieve.

SPEAKER_02 1:08:13

I love that because it it talks about how we evolve as professionals, right? Because think about it. When you thought something was difficult and then you became well versed in it, now you have the ability to open yourself up to learning something else, something new, right? Because now you you're at that level, now you're over that learning curve where you feel comfortable enough and you've you've just expanded. And isn't that how we evolve as individuals? But imagine as a professional, especially in this field, it has been an amazing conversation, Dr. Song. I want to thank you for the opportunity to be able to pick your brain and have you have you here with us today to be able to share a little bit, at least in this one hour session uh plus of uh how we can expand as professionals and as interpreters. And I would love for you to take a moment to share with our listeners how they can find out more about you and the work that you do. Where can they connect with you?

SPEAKER_00 1:09:13

Sure. Um, so I'm the founder and CEO of GreenTurp Technologies, which is a technology company dedicated uh to provide interpreting uh profession technology assistance. Um we have uh an ecosystem of tools. We call it solutions. Uh one of the solutions is uh GreenTurp's uh GD booth. It's a virtual booth with videos between interpreters and collaborations, handovers, assistance, and audio uh uh you know uh modification tools where you can modify the audio according to your needs sometimes. Um and then we also have a tool called TurpMate, which is a Kai tool, a CAI tool. Uh it's a broader sense um

Concrete Actions And Where To Connect

SPEAKER_00 1:10:00

computer assisted interpreting tool. So it has the narrow sense transcripts, real-time live transcripts. It has uh the full text translation, but it also has something more than that, like uh audio monitor where you can bring the sound from the floor uh when you're on site there uh through your computer software. And uh you also have the recording, which is a one-button recording, but it simultaneously records your interpretation and the floor voice on the edge, um, you know, in your computer. So you can replay that for your own assessment, review, uh, and improvement. And uh one special thing about our two is the models are offline, so you can safely use it without worrying about privacy or security issues or client permission issues because it doesn't go beyond your computer. It's just like any other two, like you're writing on your own piece of paper. Um, so that's something that makes a big difference. And besides, we made it completely free for interpreters to use for these two uh tools. Um, you can go to www.gtmeeting.com. That's our official website to know more about it. Feel free to reach out to me if you have any questions regarding technology for interpreters.

SPEAKER_02 1:11:23

Thank you, Dr. Song. And also please follow him if you don't on LinkedIn. Super helpful information. Dr. Song is always sharing a lot of content, a lot of things to make you just ponder, think about, pause a little bit, and think about the tools that are that are currently out there and how they are there to support us rather to than impede us. So, Dr. Bernard Song, thank you once again for the opportunity. It has been a privilege, it has been a pleasure to have you here on the show. And thank you once again.

SPEAKER_00 1:11:55

Thank you, Mira, for inviting me to this podcast. I just said this is my first time attending a podcast, but it's a really cherishable experience for me. Thank you.