Latest / Found in Interpretation Podcast / Ep .55 - Paul Appleyard on the early days of computer-assisted translation.

Transcript

Ep .55 - Paul Appleyard on the early days of computer-assisted translation. · Found in Interpretation Podcast

0:00
Hello, hello and welcome to another episode of Found in
0:05
Interpretation podcast. Once again, I'm wondering your
0:08
co-host Anabella Darwin here. Brian, Brian, how you doing?
0:11
I'm doing great. It's it's cold outside, but
0:15
other, but my heart is warm. So we're good.
0:17
And we're also here with Paul. Paul, how are you today?
0:21
I'm good, thank you. It's not as cold possibly as
0:24
where you are, it's grey. But again, I'm nice and warm and
0:27
I'm looking forward to having this trap with you.
0:29
It's gonna be great. Alright, don't forget everyone
0:32
like share, subscribe and let's kick it.
0:47
OK, so here our guests have found an interpretation podcast.
0:51
We are speaking today with Paul Appleyard.
0:55
Paul Appleyard. Actually, he has a career in
1:00
translation and interpretation, including a time, he says, in
1:06
the last century or Millennium, managing the translation process
1:13
for a number of software companies.
1:15
Paul was a member of the ITIL board for six years.
1:19
If I'm not mistaken, that's an organization in the UK for
1:23
translators and interpreters and and he was the institute's chair
1:29
at one point. He has contributed to ITIL in a
1:32
number of ways in many years, including being a network
1:36
coordinator. He has worked with a network
1:39
regional groups, he works, he worked on articles bulletin.
1:44
He's qualified in interpretation and translation.
1:47
He even won an award in 2019 for an interpreting assignment.
1:54
He's also a member of the Americans Translators
1:57
Association, SFT in France and the British Computer Society.
2:03
Paul, did I miss anything? No, it makes me sound like I'm
2:07
quite good, but it's all a lie. OK, that's sums it up.
2:12
And then out of breath after all this.
2:17
That sounds like a fascinating career though.
2:18
So you you were talking about millenniums, so I guess you
2:22
started in the previous Millennium.
2:24
As I did, yes. Gosh, that does sound like 1000
2:27
years ago, doesn't it? So how did how did you get
2:31
started? Right.
2:33
Well, I mean, depends how far back you want to go really.
2:36
I mean it's goes right back to school.
2:38
I was slightly less bad at French than I was at anything
2:41
else. So that meant essentially that I
2:46
ended up doing French at university.
2:47
And then when you're finished doing languages at university,
2:50
you've got 2 choices, which is either to teach or not to teach.
2:54
And if you choose the not to teach option, then there are
2:57
very few opportunities for you to keep using your languages.
3:01
And I was quite keen, having realized that I was, as I say,
3:04
slightly less bad than anything else.
3:06
The languages were a good way to go, so I ended up going to do
3:12
the postgraduate course at the University of Surrey, which is a
3:17
shout out to them. And then I was actually
3:22
reasonably good at computers back in, back at that time when
3:25
not everybody had a computer at home, which was a very hard
3:28
thing to believe these days. What year are we talking about?
3:31
Here we're talking around about the mid 80s now.
3:35
OHS. And because of that, after a
3:39
couple of brief assignment brief jobs that I'm sure I'll mention
3:43
at some point, I ended up working for software companies,
3:48
helping them to do the translation of their software
3:51
into different languages at different times, at different
3:54
places. Proudest boast being once to
3:57
release a software package for a US software company in 21
4:03
languages simultaneously, at a time when people just didn't do
4:06
that and this was all yeah, back in the last century.
4:10
But this is like a pre windows I'm assuming.
4:13
This is pre windows. This is.
4:16
Well, Windows 3.1 was coming out to about the same sort of time
4:21
that I'm talking about. I think I've got the disks for
4:25
it somewhere in my. How many sacks?
4:29
Oht it would be a pile about this big, all of them with about
4:33
512 K on each disk and it would take you about two days to
4:37
install it and then you'd have a problem.
4:38
And also something called OS 2 on.
4:42
There was a we. Did an IBM.
4:44
That was the IBM operating system too, because Operating
4:47
System One was their first attempt, and then Operating
4:51
System 2 came around and then they brought out OS2 Warp
4:55
because it was going to go more quickly warp speed, but it went
5:00
more quickly into oblivion. 286. Worksheet into nothingness.
5:06
On a 286 and then indeed on a 386, yeah, because you could get
5:11
the IBM PC-70 laptop, sorry, the IBM 70-PS2 pewter.
5:18
And then you've got the P70, which was the portable version
5:20
of it, which was big and heavy and really quite thick, and it
5:26
had an orange and black screen. Gosh, it does make it sound like
5:30
I've been around for. A while anyone under 40 here?
5:32
There's no idea what we're talking about.
5:33
I remember the 486 when that came out, that was big.
5:37
Like. Yeah, I had.
5:39
I started with a 386 on DOS 5.0. Yeah.
5:47
Yeah, that's that's what, that's where, that's where I started.
5:50
In the UK, I mean back at the very end of the 1970's, the
5:54
government had a really, really good idea because obviously
5:57
computers were going to be the next thing.
6:00
And somebody very, very insightful, I don't quite know
6:03
where realize this. And they came out with a program
6:06
that was, I'd have to look into the history books, but it was
6:09
run through the BBC who introduced a computer, the BBC
6:14
Micro computer, which plugged into your TV and you could
6:18
program it to play games like Pong.
6:22
Like a Commodore 60. Four, No, yes, I suppose it
6:26
probably was in a way. It was a keyboard and you could
6:28
plug it into your TV and what have you, but it, because it was
6:32
the BBC, the idea was that it was going to be educational.
6:35
Um, so it came with introductions to BASIC, the
6:40
computer language and things like that.
6:42
So people would, there was that and there was the Sinclair
6:46
computers as well, the ZX-81 and the Sinclair Spectrum, which had
6:50
colour woo. Yeah, at least four of these.
6:56
And they were instrumental in the UK at that time being quite
7:01
advanced in computer terms. Um, because the BBC would then
7:05
run TV programs are tied in. And do you remember the noises
7:08
that old modems used to make? You could actually record from
7:14
your TV that well, that's exactly right though.
7:18
But you could record those things and get some very simple
7:21
programs off your TV set by recording them and then loading
7:23
them into the BBC. Gosh.
7:27
And and that is why then companies like ARM, the chip,
7:32
the chip manufacturer, which was based in Cambridge was Acorn
7:35
Research, couldn't tell you what the M stands for, but they came
7:40
up with obviously computer chips that then developed into a
7:44
company that is now a US quoted international company doing very
7:50
nicely for itself. OK.
7:52
Yeah, You know, this is going to be a history lesson.
7:55
Oh no, I mean, I'm funny interpretation.
7:58
We have guests that when into all kinds of field and it's
8:02
really to show the people that there are million ways to become
8:06
a translator and interpreter out there, right.
8:09
You can start with computer science or whatever, whatever
8:13
suits you. So were you managing everything
8:16
the the 21 language launch and everything where you actually
8:19
doing some of the translation as well or how does that work?
8:22
No, I I managed the translation. We had a translation department
8:27
in Chicago, which had the key languages.
8:33
I'm going to have to remember how many we had in the offices
8:35
there. I think we had 7 languages in
8:37
the office in Chicago with teams of five translators in each
8:41
language working in so French, German, Spanish, Italian for
8:48
Chinese, French, Yep, Chinese and others.
8:53
And then we had other languages that were run locally by local
8:58
affiliates, local branches in the various countries.
9:01
So we had places like we had offices and check checks.
9:05
The vacuum at the time change Republic, we have Poland, we
9:09
have Russia, we had Finland, we had others, all sorts of bases.
9:13
And So what I used to do was take the computer tapes with the
9:18
translations on to various places once the translations
9:21
were done, and then I would install them into languages I
9:23
knew nothing about. The company that we were working
9:27
with then had IBM System 38, followed by our IBM S4.
9:31
Hundreds more computer acronyms from the olden days and we would
9:36
take these tapes and then later on discs and we would install
9:40
them or I would install them directly onto machines wherever.
9:44
But we're talking about a time where there was no Internet, so
9:48
you had to physically get the this from the translators in
9:51
various countries, and then you had to travel all the way to
9:54
Chicago. How did that work?
9:55
Well, I I was based in Chicago for a while, but then I was
9:59
that's rephrase that I was based in Chicago and in Europe.
10:03
And so I did, I did air miles. Let's just say that it at a time
10:09
when possibly people weren't as aware of the environmental
10:13
impact of flying as perhaps they might be now.
10:16
I was flitting backwards and forwards across the Atlantic.
10:18
And then when I wasn't flitting across the Atlantic, I was then
10:21
going into Europe. But yeah, you're absolutely
10:23
right. You were carrying physical
10:25
computer products and then going to a chewing gum factory in
10:30
Poznan or a steelworks in Catawissa or a bank in Moscow or
10:37
places like that. And then you were.
10:39
Yeah, physically posting the tape into a machine and then.
10:44
Tell me that whenever you did that, you always brought 2
10:47
copies, right? Definitely.
10:50
Or three. Of your passports?
10:52
No, just. Does everything multiple copies?
10:56
Because yes. Otherwise, if you got there and
10:58
there was a problem, then you would have to go home.
11:01
Yeah. Absolutely, Yeah.
11:03
So because I would have been quite nervous.
11:06
I mean, I think to this day, if I have a translation and it came
11:11
from someone in one country, I'm bringing it to the other
11:13
country. And then I remember those floppy
11:16
disks, if one of them crashes, then you can't.
11:19
I mean, the whole whole sequence can continue and then you would
11:23
have to go back physically, right?
11:25
Or get it by express mail or whatever the system was.
11:28
Yeah, yeah. And then we would, we would
11:30
carry with multiple copies. And if you had more than one
11:33
person travelling, each person would have a copy in their bag.
11:35
Or. But yeah.
11:37
But it was fun. I mean, you know, you, you talk
11:40
about, there's one of the things, and we still talk about
11:42
it a lot in our profession, but hard work isn't as hard if it's
11:48
fun. And that was true for me back,
11:51
Yeah, back in the last Millennium.
11:52
And it's still true today. Hard work.
11:54
That is so true. You've got to tell.
11:57
You've got to approach it in different ways.
11:59
Hard work could be stressful, and stress is not a good thing.
12:02
But if you're having fun, and you know that fun can be tapping
12:06
away on a computer and then watching a machine burst into
12:10
life with whatever you've installed working properly.
12:13
And that can give you the same satisfaction as you can then get
12:17
from interpreting a celebrity or somebody recognized and then
12:21
hearing your voice on the TV or whatever it might be.
12:24
So you can have fun. And that means that you're not
12:28
necessarily thinking about the what the effort, you're thinking
12:32
about the result and enjoying what you've done.
12:34
Absolutely. And so when did you transition
12:37
from the IT world to interpretation?
12:40
Um, well, I transitioned from interpretation into the IT world
12:45
and then back. My very, very first job had an
12:50
interpretation component. After I finished at the
12:52
University of Surrey, I went off into the world and I worked for
12:56
a freight company, a small freight company in a place
13:00
called Tonga that nobody has ever heard of, which is near a
13:02
place called Aldershot. That will mean something to
13:05
anybody watching this from the UK, but nobody else.
13:08
Aldershot was the home of the British Army, so they say on the
13:12
signs, and it was a freight company.
13:15
And at that time we used to drive trucks to various places
13:20
and across Europe. And there were meetings of
13:25
haulage companies that took place in Paris because at that
13:30
time there was no open market. So you if you had a truck from
13:36
Germany and it went to France or from the UK and it went to
13:39
France, you had to then negotiate what you were going to
13:42
bring back to your home country, which is unbelievable really.
13:48
If you think about how how many countries and how small each
13:51
country is as opposed to a larger market, it would be
13:54
insane to think that any company might voluntarily then vote to
13:58
leave such a large open market. But we won't talk about Brexit
14:01
today. Um, but if you consider this, we
14:07
used to go across to the to, to Paris and then we would
14:13
negotiate what we were going to bring back.
14:15
So I would go over to Paris in a truck and then I would change
14:19
into a suit and I would go to a meeting and I would interpreting
14:24
meetings where people were negotiating what they were
14:26
bringing back, which was again, absolutely fascinating,
14:31
absolutely eye opening. In fact, for, for a, for a young
14:36
fresh out of university chap, Um, but really, again, very
14:41
enjoyable. Sadly, it didn't last for very
14:43
long. So then I worked for a
14:44
translation company in London who would then also send me out
14:48
interpreting. Then I transitioned into it,
14:51
which I did for for about 12 years.
14:55
And then the information technology sector is very
15:00
dynamic, which is another way of saying that it changes
15:03
regularly. And when it changes regularly,
15:05
people get hired and fired. And after one occasion when I
15:09
was fired, I was given sufficient payoff to join my
15:13
wife, who was working already as a freelance translator.
15:17
Oh. And that's when I started
15:19
interpreting again properly. What what is the biggest change
15:23
of experience in the last decades?
15:28
I would say because now we are going through some changes with
15:31
AI, but do you think that this coming, this change is current
15:35
change is actually bigger or milder than everything?
15:41
Else that. You've you've experienced
15:42
before. That's a good one.
15:45
I mean, if I go again, right back to what I was telling you
15:50
before, at that time people expected software to be in
15:54
English with a few lines in other languages, the lines that
15:57
would be facing the clients. Then there was an ever
16:02
increasing expectation that people would be able to get
16:04
whatever they wanted in their own language, which is a good
16:08
thing because obviously that means that more people can be
16:10
included in whatever journey they're going.
16:12
I would say from what I've seen and the way that I've been
16:16
working with the COVID years were a sea change.
16:23
Because we're sitting here talking with our headsets on and
16:26
we're in three different countries in at least two
16:29
different time zones, many, many thousands of miles apart from
16:32
each other. And we're doing it completely
16:36
naturally. And if we had done this, what
16:38
we're doing here today, 10 years ago, we would have had a lot.
16:42
It would have been a lot more of a rustic, rudimentary process.
16:46
We wouldn't have known what we were doing.
16:47
And working online, particularly during COVID, meant that people
16:53
became aware of the fact that they could use interpreters
16:56
online for short meetings where before they wouldn't have
17:00
bothered. So, I mean, I know there are
17:04
lots of conversations about whether online is good, whether
17:07
it's the full experience, whether it's not as good as
17:09
being in the room. And I absolutely agree with
17:11
everybody who thinks that you're better off in the same room
17:13
because you're picking up nuance, you're picking up a
17:16
vibe, you're picking up things that you can't really put into
17:19
words. But it has meant that the work,
17:23
the work has become more available and you can do little
17:28
bits and pieces and you can do little bits and pieces in
17:30
different time zones for places. Again, I could travel.
17:35
I, I, I can interpret for somebody who is 4000 miles South
17:39
of here, so in the same time zone, but they would never have
17:42
flown me 4000 miles South of here for a three day meeting,
17:46
Sorry, the three hour meeting. But now I can do that.
17:49
And that means that they're getting whatever meeting in
17:52
whichever language they want to hear it for however many
17:56
languages plural that they want to hear it in.
17:58
So I would say that that was a real change in the way that we
18:05
do things, which presaged the coming AI change as well, but
18:11
was probably even more of a sea change in many ways than than
18:15
than what we're seeing now. Well, I think that it's it gives
18:22
us flexibility and choice. We don't have to live like I'm
18:27
in Quebec City and I have to admit that if technology wasn't
18:32
around, I'm not sure I could rely only on interpretation,
18:36
just like I'm doing now, or I would have had to move to, let's
18:40
say, Montreal, Ottawa, Toronto, or a much bigger place in order
18:44
to make a living. But now I can live in a small
18:48
town and yeah, Quebec is small and and I have flexibility and
18:55
choice and it gives us more work at the same time.
18:58
But I understand that some interpreters really miss the in
19:02
person setting. I do as well a lot.
19:05
But I think that this flexibility actually has allowed
19:10
us to enhance our quality of life.
19:13
Yeah, absolutely. There's no more of that.
19:15
Chasing for the last flight on a Friday night and then getting
19:19
home really late. Yeah, that's a, that's a
19:22
definite positive. I found one of the big changes,
19:24
um, before that was maybe 2004, 2005, when like more people were
19:31
having bringing their laptops into the booth and using more
19:36
terminology instead of like bringing books, like
19:39
encyclopedias and stuff into the booth like you're doing a
19:42
medical conference or bring your medical glossaries and
19:47
everything. Like when everything got
19:48
digitalized. I think that was a huge change
19:51
for us too. That's very, very true, yeah.
19:54
Although of course there was. That had its challenges.
19:57
Those first laptops had really noisy fans and could really
20:01
throw out a lot of heat. Were they warming up the booth
20:04
so they wouldn't Warmer than it is now?
20:06
Yeah. Yep, Yep.
20:08
Yeah, sauna in there. Yeah, the Scandinavians would
20:11
have loved it. That's true.
20:13
And yet you have these up in the ceiling.
20:15
You'd have these little holes and sometimes they'd be a little
20:18
electric fan, little bit like the fan on the back of a desktop
20:21
computer, but without the power running it.
20:24
So you'd have this little thing turning around, Yes, Well.
20:28
That's our noise, yeah, like the fans and the from the computers
20:32
and from the booth and like, yeah, I don't know how those
20:35
microphones to catch our voices without.
20:39
And what about other changes, let's say cultural changes over
20:43
time? Do you think that the content of
20:46
the conferences has changed? Because in my case, I've been at
20:50
about it for about 8 years now and I I started to see a shift
20:56
pre and post COVID in the content of some of the
21:01
conferences that I'm doing. I do a lot more human oriented.
21:07
I would call it conferences then I used to like before it was
21:13
really OK, the mining industry and you had something very
21:16
technical. I had that Cranberries, it was
21:18
very technical. You know, you had very technical
21:21
conferences. But now with a lot of cultural
21:25
shifts, I think that people don't see interpretation in the
21:29
same way. Have you noticed that as well?
21:31
I know exactly what you mean about the the human context,
21:35
that human content. Let me say that properly, you're
21:39
absolutely right. And a lot of what I do is is
21:42
focused on making lives better in whatever way it might be, be
21:47
it through farming. It can be a farming conference,
21:49
which will be a farming conference about farming better,
21:53
or it'll be a government conference about making lives
21:58
better. I've done two jobs this week
22:03
just on improving healthcare outcomes for people.
22:06
But you're absolutely right, there's a focus on people in all
22:09
of those. Um, back in the back in the day,
22:13
I used to do a lot of work for one particular company in the
22:19
wider defence sector where humans weren't mentioned at all.
22:23
And anything that I hear, I don't do much for them now, but
22:26
anything that I do hear about will again, focus on the human
22:29
aspect. So it's a good sweeping
22:32
generalisation an absolutely true.
22:33
Yeah, I found a lot of conferences now it's like more
22:36
like work life balance and DI like a lot.
22:41
Yeah, a lot more human aspects like the diversity, equity,
22:47
inclusion and, and accommodations and, and that
22:52
kind of thing. So that's, yeah, I've noticed a
22:54
huge shift in that. Like I remember back in the
22:57
early like the 2000s, like, yeah, we were doing the how to
23:03
global positioning systems for satellites so they wouldn't
23:05
interfere with one another and that kind of thing.
23:07
And now we're like, more like, how do we work together better?
23:12
Yeah. And ultimately, I mean, that's
23:15
driven, it's driven by companies who are looking to achieve value
23:20
for their shareholders. But isn't it outside government?
23:23
Everything is about ensuring that whoever is buying your
23:26
services is generating value for whichever market therein, or
23:30
whichever shareholder they want to keep happy or what have you.
23:33
And if they've realised that they can do that by having
23:36
better outcomes for the people working for them or the people
23:38
supplying them or buying from them, then I think we all win.
23:43
Is the world becoming more ethical in a way?
23:48
Yes and no. The pencil you ask, I guess.
23:52
It depends who you ask. I would say it is, uh, inasmuch
23:56
as everything we've been saying about people worrying about the
24:00
human side of things. But I would also say no,
24:03
because, um, we can move on to AI if we want to.
24:09
OHS we, we, we've scratched the surface, but please, let's
24:12
continue. I mean, there is a lot of talk
24:16
AI. Gosh, what goes straight in
24:19
then? OK, AI is a tool like any other
24:22
tool. You know, I'm talking about when
24:23
I used to take computer tapes and people would expect things
24:25
in their own languages. It's all that rubbish.
24:29
AI is a tool like that, but it is not yet perceived as a tool
24:33
because we're still at that happy point where it's a
24:35
wonderful, wonderful solution. It's going to cover our plates
24:40
in freshly mined gold and we're all going to be extremely rich
24:45
except for all the people who aren't going to be extremely
24:48
rich because they're going to be left one side in the pursuit of
24:51
this El Dorado. So I'm going to go off now as I
25:00
was wander off and I've lost my thread now.
25:03
Just started and then. Go please OHK Well, I, I the way
25:09
I see it is some I was watching another podcast a couple of days
25:14
ago and there was the discussion to see is AI making people
25:21
smarter or Dumber? And then someone was saying that
25:25
it makes, you know, people who have less abilities in certain
25:30
ways, like to write a text or to write a text without mistakes.
25:34
It would basically level up. But the problem is people who
25:38
are educated, you can't tell the difference anymore.
25:42
And then I'm more of the opinion that it's that aspect is true.
25:46
But for those who are able to interact with the device,
25:52
because AI is a device, it's a tool, will be able to get even
25:57
higher because they will be able to maximize and come up some
26:02
with some creative ways to use it.
26:04
So yes, everyone will be grammar and mistake free, but at the
26:11
same time, that'll be it. So you'll have to get your game
26:15
even higher if you want not to be forgotten in this changing
26:20
economy. I think you're absolutely right
26:23
with that and that again, it comes back to the fact that it
26:26
is a tool. You know, if if you if you rode
26:29
your first bicycle when you're a child, it didn't have gears and
26:32
you fell off, then you learn how not to fall off.
26:34
Then you got a bigger bicycle, which possibly had three gears.
26:37
Then you got a bicycle with five is.
26:38
Then you got one with 55 gears, whatever it might be.
26:42
And every time you've learned how to use it without falling
26:45
off and you've gone faster as a result of it.
26:48
You're absolutely right. It is just like that.
26:50
But one thing that I don't think that we're yet focusing on all
26:54
that much is the people who don't necessarily managed to
26:58
keep up through no fault of their own.
27:01
And I think we're seeing that on the translation side rather than
27:04
than the interpretation side because the people who you're
27:08
mentioning who can now write grammar error free letters
27:14
without necessarily saying other than anything other than OK,
27:18
ChatGPT, write me a letter to complain about whatever.
27:22
Um, the translators who are being left to one side because
27:29
people are now saying, OK, tragedy, BPD, translate this
27:32
letter into French for me, haven't necessarily been
27:36
considered as we move towards going forwards because those
27:39
people aren't necessarily getting an opportunity to just
27:42
use AI as a tool in their work because their work is being
27:45
removed from them. And we haven't necessarily had
27:50
an honest discussion about those people who are being left to one
27:53
side through no fault of their own.
27:55
So yes, it is a tool and it is a tool that we can all use, but
27:59
within a wider society we need to think about the people who
28:02
don't get the opportunity to use that tool.
28:06
That's a bit of a manifesto. Have you, has your wife
28:10
experienced that? Because you were mentioning that
28:12
your wife is a translator, right?
28:14
Yeah, she she has. Yes, she has had a clients who
28:19
have just disappeared. She has had clients who have
28:23
moved on to the post editing machine translation model where
28:28
the rates being offered to her are lower.
28:33
The corollary of that is that theoretically you can post edit
28:37
10,000 words in a day. So theoretically you are not
28:41
necessarily losing out if that's what you're doing, but it's not
28:45
many translators. Again, I'm talking about
28:47
translators we can set to interpreters in the future as
28:50
well. But many translators, I think
28:53
are quite creative in the way that they approach their work
28:55
and and in what they do. Even if it's a boring manual of
29:00
some sort, they are approaching it with an in a creative way
29:04
that they cannot get if all they're actually doing is
29:07
looking at computer produced text.
29:10
Well, but I have a question about that because what is
29:15
exactly? I'm, I'm, I'm an interpreter.
29:18
I have a bachelor's degree in translation but I haven't really
29:20
worked as a translator in like the last decade or so.
29:25
And when we talk about post editing it must be more than
29:32
let's say copy pasted to ChatGPT.
29:35
Give the text to the the translator and say just correct
29:38
that. Because the translator.
29:39
The translator. Can do that too right?
29:42
It's like a 5 minute thing. So is that it?
29:45
Is that what post editing? Is to a certain extent, yes.
29:48
I mean, again, it depends on the clients and what have you.
29:52
So one of her clients, she has done the translation for them
29:56
for literally a quarter of a century.
29:59
Um, so they have got a massive translation memory in trados or
30:04
whatever that theoretically has all of the right terminology up
30:09
to the point when they decided to move on to a post editing
30:11
model. So theoretically she would be
30:15
scanning through the text and she would be recognizing things
30:17
that she might have translated 10 years ago, five years ago and
30:21
what have you. And that would mean that the
30:24
output using possibly an in house, an in house tool is going
30:31
to be reasonably accurate. But overtime it's not
30:35
necessarily going to remain as accurate because people are
30:37
going to continue to write new texts in English or possibly
30:40
generate new text English, which they would then have translated
30:43
into other languages. But if it can't align with
30:46
something that's already in a database, because any
30:50
translation memory of any sort, whatever you're talking about,
30:52
even an AI is effectively a database full of words or terms
30:56
or so the person doing the post editing in the case of someone
31:01
like my wife, is somebody who has a lot of knowledge in a
31:04
particular subject area for a particular client.
31:09
But you're not necessarily going to get longer term guarantees
31:12
about the consistent quality that you would get from using
31:15
the same human using a tool like Trados to ensure that their work
31:20
is consistent, because an AI isn't necessarily going to be
31:24
aligning it in the same way. Now in the the shift that you
31:28
explained for translation with your wife, are you seeing a in
31:33
contrast with interpretation? Have you seen like a decrease in
31:35
business and like with your UK clients or international
31:38
clients, like more people that are trying to use like like
31:43
speech translation or something like that?
31:46
I think I have done a couple of jobs where AI has been used as
31:51
well or has been tested as well. And clearly there are going to
31:55
be certain contexts in which it's going to work extremely
31:58
well. But quite a bit of the work that
32:01
I do possibly like, like you as well, is in context where there
32:06
are multiple languages. You've got a language being
32:08
spoken on the floor and then running through multiple
32:11
languages in booths, be they virtual or real, and often in
32:16
situations where there's more than one sound coming in at
32:19
once. And an AI is not yet at a point
32:23
where it can distinguish background noise and clear
32:27
foreground noise and then come up with something that can then
32:29
multiple, that can then move through the multiple languages
32:33
where I've read the AI is already being used.
32:38
And in fact, one company I work for is using it in that context,
32:40
is in a medical context in the US where they are using it for a
32:48
one to one delivery of an interpretation where previously
32:54
it might have been an unqualified family member
32:56
working with it or where the terminology is very clear.
32:59
And it could do it in a way which I suspect it's probably
33:03
much better than going through a non qualified medical
33:06
interpreter who shouldn't be telling somebody that their legs
33:10
are about to be chopped off or they're going to be dead next
33:12
week or in the in the worst of context.
33:15
So, but no, it's certainly in the work that I do, I have not
33:19
yet seen an impact whether it will come.
33:24
You need wiser people than me who have a better overview of
33:27
the whole thing to be able to say whether it will.
33:29
I suspect that there will be attempts.
33:32
Yeah. Well as they already there as as
33:34
a business owner would you ever think of offering that as a
33:37
solution for your clients like the B say the operator of the
33:42
AI? I would like to say that given
33:46
where I am in my life and in my business and in my approach to
33:50
things, if there is a difference to be made between using an AI
33:55
and not using an AI, I would like to market myself as the
33:57
human approach things. Because there is a lot of, there
34:04
is a lot which you can do with machines, but there is also
34:07
something that you get from sitting in the same room as
34:10
somebody. Sometimes you can, you can, I
34:13
don't know what we call it. You call it a nice warm feeling,
34:16
which doesn't you? You can't put that into a
34:20
computer context, but just sitting there with a human being
34:24
sometimes is going to be something that you can market as
34:28
a plus. And I think that the more the
34:30
machines come along, the more the human side can be marketed
34:35
as a plus. Not denigrating the machines
34:38
because the market as a whole, I suspect, is going to get wider.
34:41
Again, you go back to what I said right back, and I said it
34:45
two or three Times Now when people expected a few lines to
34:48
be translated and that was that people are going to want more
34:51
and more and more and people are going to interact more and more
34:54
and more. I suspect a cross borders and
34:59
there's going to be a there is going to be a place for the
35:01
human and I would rather offer the human than the computer,
35:04
despite being a member of the computer society.
35:07
Well, I always, whenever people fear a technological shifts, I
35:11
always say that human needs are infinite.
35:14
That's the only thing that is infinite.
35:16
No, it seems we found creative ways of always wanting more.
35:21
So I think that's the reason why technological shifts, usually
35:28
we'll see about this one, but usually in the end brings about
35:32
more opportunities and more work than we expect.
35:37
Because people, I think start from the frame where this is
35:41
what it is now and technology will replace me, but they don't
35:45
think of the gazillion possibilities that will be
35:50
created out there that will make some shifts and then it will
35:54
give us even more opportunities. I, I, I'm reminded always of
35:59
this interpreter friend that was always concerned about, you
36:05
know, the future and where things are going a few years
36:08
ago. And, and even I met him before
36:12
COVID. And then about two years ago, I
36:15
said, hey, by the way, we'll mention his name.
36:19
He might recognize himself. But I said, hey, do you think
36:22
that you have more work or less worse than you expected five
36:25
years ago when you were basically concerned about the
36:29
future? And he was like, yeah, I didn't
36:31
expect that. I said, yeah, exactly.
36:33
That's it. The the proof is in the pudding.
36:36
Yeah, we don't have a crystal ball, that's for sure.
36:38
That's very true. Now is ATITI doing anything with
36:43
AI or like doing more promotion of human interpretation or are
36:49
they playing a role in in this whole shift?
36:53
Well, I mean, I'm only a member now, I haven't been on the board
36:55
for a few years, but I TI is, I would say, doing quite a good
37:00
job of working to find the balance.
37:04
They're working with again, the universities, University of
37:07
Surrey on courses on AI which and other webinars and seminars
37:13
about AI. They are offering quite a few
37:17
webinars in fact, on, on the business of translation, the
37:20
business of interpreting, Um, but a but I TI is first and
37:25
foremost a member organisation. So it has to respond to the
37:29
needs all of its members and the expectations of its members.
37:33
That, that was true when I was chair.
37:35
It's true 2 chairs on from me. The current chair, I think has
37:39
the same focus on, on the membership of, of that
37:45
organization. I TI was founded 39 years ago.
37:49
So next year is their 40th year, and their key mission has always
37:54
been to focus solely on the professions of translation and
37:58
interpreting. Back when they started, people
38:01
didn't have WordPress as necessarily.
38:05
People used to send tapes down to the translators used to send
38:08
tapes to the to the typing pool and then they would be typed up
38:11
by the girls in the typing pool and then sent back U to them.
38:15
Or you would print stuff out and send it by fax.
38:18
Yeah, yeah. Or even telex.
38:20
Are they regulated by the government, just like here in
38:23
Quebec? Is no not.
38:29
I don't believe there are regulated like OTAC.
38:31
There is. The Chartered Institute of
38:35
Linguists has a Royal Charter, which is the other large
38:40
linguists organization in the UK, which has been around for
38:48
longer as well, but is focused on language in a wider sense.
38:52
So you'll have teachers, people who are interested in languages,
38:56
but there is not. I mean, it's an unregulated
38:59
profession. Effectively, you know, anybody
39:01
could. My neighbor could set themselves
39:04
up and put an advert in the Yellow Pages and say I am an
39:06
interpreter and nobody would notice until they turned out for
39:09
a job and realized they couldn't speak.
39:13
So they're not regular, they're not strictly regulated.
39:16
They are. There are obviously regulations
39:17
around membership and being in the not not-for-profit sector
39:21
and what have you, but the proof of their existence is in the
39:25
pudding of their existence I would say.
39:28
But I'm talking as I say. I am talking as a humble member
39:31
now, having done my time. And what do you expect for the
39:36
future? Well, one thing I would say is
39:40
that, and again, that's something I said quite a long
39:44
time ago, we do something that is fun.
39:49
And I would like to think that the future, in whatever form it
39:55
takes, will hopefully still be fun.
39:59
It may well be that we are sitting working at home
40:02
uniquely. We don't go out anymore.
40:05
It may well be that there is a slight shift to having the human
40:08
in the room, which again, I've alluded to before.
40:11
But I would say that the future for interpretation is probably
40:14
bright. I would say that the
40:16
expectations are that people are going to continue to talk to
40:19
each other in whatever context. And I would like to think that
40:23
I'm going to be along for the next part of the journey.
40:26
Yeah, no, I remember even like in 2008 when there was the big
40:29
global economic crisis, we still had to interpret all the people
40:34
who were trying to get out of the big mess that was created.
40:37
So like, people meet when there's in good times and in bad
40:41
times. People always need to talk.
40:43
People needed to talk through a pandemic.
40:45
I mean, even if it wasn't in the same room.
40:48
So, yeah, I think that we're, I think that the pandemic in a way
40:52
has created a foolproof system for people to continue to
40:56
communicate through many years to come.
40:59
Very, very true. Well, until people put little
41:03
devices in their head and communicate with their thoughts,
41:05
I think we'll be around for awhile.
41:07
I don't want that. I don't want my.
41:10
I don't want my thoughts unfiltered being broadcast
41:13
everywhere. It's already hard to filter what
41:16
I'm saying here out loud. All right, Paul, thank you very
41:24
much for participating in the podcast, Brian.
41:28
Yes, and thank you so much for bringing all that knowledge and
41:31
like the, the, the transition from one Millennium to the next.
41:38
Every time you say it, I sound older.
41:42
Well, I'm sure you're gonna continue and be there at least
41:46
for the next decade as well. Most definitely.
41:49
Thank you very much. Don't forget, like, share,
41:53
subscribe and we'll meet each other for another episode next
41:57
time. Thank you everyone.