ITI Conference 2024

“New world, new work” was the theme of the ITI Conference 2024, held at the John McIntyre Conference Centre in Edinburgh (and online) on 4 and 5 June. The event featured engaging speakers, thought-provoking keynotes, and useful sessions, many of which explored the challenges in the new world of work where artificial intelligence (AI) increasingly encroaches upon the territory of translators and interpreters.

 

What does the future hold for translators and interpreters
in an AI-dominated world?

Mingling with like-minded professionals is always a delight, and I felt warmly welcomed by everyone I encountered. ITI conferences are renowned for being vibrant events where practical knowledge is shared, generous advice is given, and inspiring ideas are conveyed in an open, convivial and friendly atmosphere.

The urgent need to adapt to AI

My motivation for attending was twofold: meeting or reconnecting with fellow professionals and friends within ITI as well as comparing notes with them; and seeking guidance on how to brace myself for the new AI world. I translate patents and so far have not yet had to “post-edit” a machine translation myself. I can’t help wondering for how much longer I’ll actually be able to carry on working in my bubble.

 

A key message conveyed at the conference was the urgent need to adapt our skills to the far reaching challenges posed by AI. AI is already demonstrating remarkable capabilities, and I have no doubt it will continue to improve at mind-boggling speed. What does the future hold for me and other translators as well as interpreters? 

As Sara Robertson, ITI’s new Chief Executive, pointed out in her keynote speech, we need to increase our ability and willingness to be resilient, adaptable, entrepreneurial, our own marketing team, the solution to our clients’ problems, and professionals. She proposed that if the theme of the conference is “New world, new work”, then perhaps there also needs to be a new you!

 

To compete with AI, we will need to be resilient, adaptable, entrepreneurial,
our own marketing team, the solution to our clients’ problems, and professionals

 

A tool for generating ideas and first drafts

It was suggested in several sessions that AI is just a tool. It can therefore, for example, act as an idea generator in SEO translation, as mentioned by Tess Whitty MITI in her talk “Adding SEO expertise to compete with AI”. According to Bex Elder MITI in her talk “Raging against the machine: Is translation really dying?”, AI could serve as a helpful starting point in translation. Terence Lewis MITI, in his conversation with Cate Avery FITI on the topic “AI: Will there be a place for me?”, pointed out that AI output can be considered a first draft and could be useful for brainstorming.

Terence predicted that the first stage of the translation process will eventually be taken over by AI, but he also stressed that AI can easily get the wrong end of the stick. After all, AI cannot understand people, languages, and cultures; it just understands numbers. AI cannot read between the lines, which is so important in translation and interpreting. While we must absolutely acknowledge the benefits of AI, we should also make clients aware of the risks involved in using it.

 

The ITI Conference 2024 was held in Edinburgh
on 4 and 5 June 2024

Your accent defines who you are

The conference programme was varied and so vast that reporting in detail on individual talks would go beyond the scope of this blog article. I’m therefore sharing just a few of my own impressions and gleanings. Note that the points mentioned here were not necessarily the key points in the relevant session, but stood out for me personally.

 

Do interpreters’ accents matter? I’m not an interpreter, but the title of Katherine Dagleish’s session appealed to me instantly. Since I’ve become (perhaps excessively) self-conscious about my accent, it was interesting to learn that even native speakers can feel insecure because of their native accents. Katherine reported that, according to a poll, 10% of adults felt teased because they didn’t have “the right accent”, while 17% felt they had missed out on jobs because of it. It is important to realise – to quote from one of Katherine’s slides – that your accent defines who you are and can also locate you economically and socially.

My most enjoyable ITI conference ever

I would like to extend huge thanks to the organisers for putting on this wonderful and beneficial event and also for the lovely bamboo coaster! By the way, it spoke to my minimalist heart that on registration, in addition to my name badge, I wasn’t handed the usual bag full of leaflets and freebies this time. The only real niggle I can think of was the use of cheap plastic plates, glasses and cups in the cafeteria during lunch, which seemed slightly out of place in the otherwise sophisticated environment of the conference centre.

The John McIntyre Conference Centre was a perfect venue

The John McIntyre Conference Centre, with its stunning backdrop of Holyrood Park and Arthur’s Seat in the charming city of Edinburgh, was, in my opinion, a perfect location. Many of us stayed in the student accommodation a very short distance from the venue.

For me, this felt like the most enjoyable ITI conference ever, although I can’t quite put my finger on why. Perhaps it’s because the “less grand” venue of the John McIntyre Conference Centre felt more comfortable. Or perhaps it’s simply because I have finally learnt how to manage my energy a bit better during a large-scale event such as this one. 

Uneasiness about AI and a sense of positivity

All in all, the ITI Conference has once again provided me with the opportunity to learn, network and evolve a bit more as a professional translator. I am still feeling uneasy about our new world of work in which AI is becoming ubiquitous, but I have come away with a renewed sense of positivity and confidence about what I do and what might lie ahead.

 

Mingling with like-minded professionals at ITI conferences is always a delight
(image courtesy of ITI)

Why translators don’t fear the machines

The takeover of translations by machines is impending (or so we've been told).Why then don’t human translators fear the much talked-about rise of the machines?

As I see it, it all boils down to one simple answer: translators don’t fear the machines because a translation is created in a series of stages.


Most translations require human input

Machine translation is sometimes helpful in the first stage of creating a translation, but it then cannot contribute to what happens in subsequent stages. And where machine translation is no longer helpful, a human translator’s input will be required. 

Why don't human translators fear the rise of the machines?
(Image source: Peggy and Marco Lachmann-Anke on Pixabay)

The translation stages where machine translation is not helpful include, for example:

- Researching terminology in the particular field of the text

- Identifying and pointing out issues in the source text to the client, using appropriate grammatical terminology to describe and explain those issues, suggesting improvements

- Discussing the approach to “untranslatable” terms with the client

- Finding workaround solutions to tricky terms and phrases

- Applying client style guidelines to the translation

- Creating coherence between the individual parts of the text

- Improving the first draft of a translation (also known as “rough translation”)

- Improving the translation further

- Checking that correct punctuation has been used

- Formatting the file

- Eradicating errors (including errors potentially introduced by machine translation!)

- Printing off the translation and checking it on paper

- Double-checking that correct numbers and/or reference numerals (in patents) have been used

- Rewriting the translation (where required) so that it reads like a text that is idiomatically phrased in the target language

- Ensuring that the underlying meaning of the original text has been accurately conveyed (as we know, language is full of ambiguities!)

- Checking that technical terms have been used consistently throughout the translation

- Editing, fine-tuning and polishing the translated text

- Putting a human touch to the translation


Anyone who believes that a translation can be produced by the simple push of a button is unaware that a translation is created in stages. Machine translation may be useful during the first of those stages, but creating a fit-for-purpose translation is a long, drawn-out and intricate process.

A good translation cannot be produced by the simple push of a button
(Image source: Gerd Altmann on Pixabay)

Afterthought: Nobody knows, of course, what's still going to happen on the AI front, and some of the tasks above will maybe be taken over by robots one day. Right now, we're still very far away from it. I also personally believe that we will never get to a stage where robots will be like humans.


(A German translation of this blog article is available here.) 

Machine translation in human translation workflows

With the cognitive computing age approaching at mind-boggling speed (before humans and technology likely will merge from about 2040), there seems to be a certain urgency in the need to familiarise ourselves with Artificial Intelligence. For translators this involves thinking about how (and whether!) to integrate machine translation into their workflows.

Post-editing a translation is not the same as revising it!

On 24 January 2018 an event on the use of machine translation in professional contexts was held at Clifton Hill House in Bristol. It had been organised by the University of Bristol in partnership with Universidad Pablo de Olavide in Seville and the ITI Western Regional Group (WRG), attracting academics, professional translators, translation companies and technology providers.

My main takeaways from the event:

The job of post-editor is a relatively new profession. Post-editing nowadays is either offered as a service in its own right or just used as a tool that is incorporated into the translation process.

Post-editing has been defined in the ISO 18587 standard. Yet, although it’s been defined and hence should be clear-cut, in practice it’s more complicated since clients tend to have different requirements.

Final questions, from @lnunesvieira:
1) Are more MT-friendly pricing mechanisms needed?
2) Do we need to train translators differently for post-editing?
3) What about the 'human side' of quality?
4) How to ensure positive use of activity tracking?#MTBristol

— Bristol MA Trans (@BristolMATrans) 24. Januar 2018

Machine translations often are over-edited, rather than under-edited. It is therefore important to note that post-editing a translation is not the same as revising it! They are two different skills.

Ideally, MT should be regarded as an additional tool, or translation memory, or source of reference, which for certain projects (!) can help improve efficiency and productivity.

A useful reminder from Rosario of @tatutrad: you can have good work that's cheap; cheap work that's fast; fast work that's good; but you can't have fast, cheap and good. You have to pick two!#MTBristol pic.twitter.com/BvifeHQI03

— Carol O'Sullivan (@carol_osullivan) January 24, 2018

There will inevitably need to be a move from word count-based pricing to time-based pricing for projects involving the post-editing of machine translations.

There has been a notable shift in the perception towards MT among translators because it’s becoming more capable of producing results that are usable. However, feelings of uneasiness, or strong dislike, towards MT continue to persist.


News headlines about advances in machine translation have led to inflated expectations by clients of what such tools can do. It’s worth bearing in mind we’re still very far from the point where machines can take over from us!

The upside of such news headlines, on the other hand, is they’ve drawn attention to professional translation and interpreting, an industry which had previously often been overlooked.

DeepL: Tool or threat for translators?

The end of August saw the launch of DeepL, a new machine translation tool developed by Cologne-based start-up DeepL GmbH (formerly Linguee GmbH). It was born from Linguee, a translation tool that has been around for some years and is a popular resource amongst us translators.

DeepL apparently performs better than any of its rivals’ products because it’s based on the relatively new Neural Machine Translation (NMT) approach, in which the processing of data is modelled on thought processes as they occur in the human brain. Its makers also claim to have created one of the world’s most powerful supercomputers, conveniently located in Iceland (where electricity costs are lower than elsewhere).

Neural Machine Translation (NMT) is modelled on thought processes in the human brain

Curious about these latest developments in machine translation (MT), I incorporated DeepL into my own work last week so I could familiarise myself with it. Since I’d heard it supposedly is excellent at what it does, I started off my experiment with a bit of a feeling of dread in my stomach. I was soon relieved, though, when I realised it’s basically yet just another tool. However, unlike many of its predecessors, it produces some output that is actually usable!

Having said that, I also encountered severe (in some text types potentially even dangerous!) issues in the DeepL MT output. They may seem minor or insignificant if you don’t work with language professionally; yet in translation for the commercial world they do matter. They do, in fact, matter very much!

I’m going to list a handful of these issues from the patent I was translating assisted by DeepL. (Note that for this article I’ve deliberately picked just shorter sentences or terms from shorter sentences, as DeepL couldn’t cope with longer sentences or shorter sentences with more convoluted syntax.)

“In one embodiment, the guide tube 106 includes an opening 105 on a first end which receives the medications.”
Although I was supplied with a sentence in perfect German grammar, so at first sight there seemed nothing wrong with it, DeepL had incorrectly assumed that the relative pronoun refers back to “a first end”, whereas its actual antecedent is “an opening”.

“treatment of the surface of the guide tube 106 that comes in contact with the pill”
Here we have the same issue as above: The antecedent of the relative pronoun “that” in this particular context is “surface”, i.e. not “guide tube”, because the surface comes into contact with the pill. How can a computer decide what the antecedent of a relative pronoun is? It can’t.

“The shape of the guide tube 106, the orientation of the guide tube 106 to the force of gravity or other source of force, and the coefficients of friction and drag can be specifically designed to orient the axis of each pill in the direction of travel or with the axis of the tube 106.”
“direction of travel” was nonsensically translated by DeepL as “Fahrtrichtung”, which would, of course, be the correct term in automotive contexts, whereas here it simply means the pill is moving in a particular direction.

“ridges”
Translated by DeepL as “Rillen”. Further down in the text, though, and especially when I looked at the technical drawings, it became clear that “Erhöhungen” or a synonymous term is more appropriate because the ridges on the internal (i.e. not the external) surface are described.

“low-distortion transparent material”
Translated by DeepL as “verzerrungsarmes transparentes Material”, which does not make sense here since “low-distortion” in this particular context simply means the material in question isn’t prone to becoming deformed. (Also, DeepL omitted the important comma between the two adjectives in German.)

“cameras with fast shutters”
Translated by DeepL as “Kameras mit schnellen Shuttern”; however, people working in this field tend to call them “Ultrakurzzeitkameras”.

“System 700 includes an image analyzer 704 and includes or has access to an image database 706.”
Translated by DeepL as “Das System 700 verfügt über einen Bildanalysator 704 und eine Bilddatenbank 706”. Although the sentence is correct grammatically and sort of conveys the meaning, leaving out parts of a sentence is a no-go, especially in patent translation.

“In one embodiment, the light sources are continuous.”
Translated by DeepL as “In einer Ausführungsform sind die Lichtquellen durchgehend”. The grammar is impeccable, yet the sentence sounds odd. A human translator would likely opt for a more technically sounding translation such as “In einer Ausführungsform sind die Lichtquellen Dauerlichtquellen.”

“optics”
Translated by DeepL by “Optiken” in the plural. Difficult for a computer to get right, but Germans tend to use the term in the singular here to refer to an assembly of optical elements.

“electrophoresis (e.g., capillary)”
Translated by DeepL as “Elektrophorese (z. B. Kapillare)”. A human translator would likely elaborate a bit and render the whole phrase as “Elektrophorese (z. B. Kapillarelektrophorese)” as otherwise it all somehow doesn’t fit together.

“limit the invention to the precise forms disclosed”
“forms” was translated by DeepL literally as “Formen”. In this particular sentence, however, its meaning in the patent is “embodiments” or “forms of embodiment”, so it really should have been output as “Ausführungsformen” (or “Ausführungen”, which is even more common in patents originally drafted in German).

Following my experiment, I can confirm DeepL is indeed more precise and nuanced than any of the other machine translations that I’ve previously seen floating around the internet. So should we translators see DeepL as a threat? Will it disrupt the translation industry? I don’t believe it will. Machine translation is becoming more and more widespread, but: I am convinced human input will always be required for many text types.

For any change that looks potentially disruptive, there is both threat and opportunity. It’s ultimately all about how we respond to such changes! It’s also worth remembering there is a shortage of translators (read: good translators) across the board, while translation volumes are increasing year by year. So there is no other way than additionally employing machine translation for all the easier-to-handle-texts that require to be translated.

Machine translation or MT (also often referred to as instant, automated or automatic translation) was pioneered in the 1950s, and although this has taken a very long time, machines are gradually becoming better at translating. We have to acknowledge they are now no longer producing the hopeless gibberish of the early days of MT.

I have until recently been sceptical about the viability of post-editing machine translations as a new field of work in professional translation, simply because the MT output has typically been poor. But following these latest developments, I wonder if it is now worth exploring a bit more? Although DeepL hasn’t set out its vision yet, I wouldn’t mind if DeepL was made available for professionals at some stage – perhaps as a plug-in in the CAT software that we use?


If computers are indeed becoming more and more capable of taking over the boring bits of our work, then this can only be a welcome move forward. For it’ll mean we will at last be able to concentrate and spend more time on the bits in our texts that are actually interesting, that are blissfully complex and therefore worth getting our teeth into!