The elephant in my room…

“In the style of Dali: the elephant in the room, sitting at the boardroom table discussing artificial intelligence.” DALL·EThe reaction I rarely see when discussing artificial intelligence with anyone is indifference.  The reactions I usually see are split between overflowing enthusiasm and overflowing concern.  I rarely have a conversation about them both.  But after writing a few articles on how useful it is, and obviously I spend most of my time in the overflowing with enthusiasm camp, I wanted to address the elephant in the room.

The rapid development of AI presents many wide ranging challenges… for example addressing algorithmic bias and discrimination, developing ethical AI frameworks, managing job displacement and workforce transitions, promoting AI-focused education and reskilling initiatives, ensuring AI transparency and explainability, addressing data privacy, prioritising safety research in AGI (Artificial General Intelligence) development, creating regulatory frameworks and policies etc.

AI technologies raise concerns about the concentration of power and wealth, monopolies, and their potential to exacerbate inequalities.  Its potential role in worker exploitation and political manipulation highlights the need for comprehensive regulation and policy development.  If you’ve been listening to Bernie Sanders as he does the rounds marketing his new book “It’s OK to Be Angry About Capitalism” I think we certainly do have reason to be concerned.  AI wasn’t needed to create the economic inequality we see today… the elitism of the privileged few has done a pretty good job without it… but it could make things even harder to address than they are now.  A recent article in the Guardian addressed the elephant as Sam Altman, the CEO of OpenAI which is the company behind the development of ChatGPT, was quoted as saying “we’ve got to be careful here”, and “we are a little bit scared of this”.  So whilst I don’t want to be seen as a conspiracy theorist, and while I want to look at the opportunities that could be created from technology like this, I do believe that the behaviour of the few who hold so much influence over our world today show us that we do have plenty to be scared about.

But there is little point in just being scared.  We can’t ignore the inevitable advancement of this technology or we’ll only see it being used for the benefit of the few.  We need to embrace the good and the bad so that we are able to openly address the challenges it will bring, and work together to help ensure that these AI technologies are not abused for the benefit of a select few and to ensure a more equitable future.

my elephant…

I titled this article “The elephant in my room…” because I see my part in all of this in two ways:

  1. supporting the wider initiatives to ensure an equitable world wherever I can, and
  2. actively looking at how myself, my family, my colleagues in the localization space, can embrace the future in a world that is increasingly dominated by technology

For this article I’m focusing specifically on my industry, and with the RWS Campus event on the horizon this is indeed very topical as we have so many young people entering our profession every year.  We’ve already been through a quite rapid change with the introduction of machine translation where I can remember not so long ago so many translators posting into the public forums exclaiming how useless it was and that a professional translator would never use it.  Come forward a few years and the posts we see when machine translation isn’t available would make you think the world had fallen in!  RWS published their Translation Technology Insights 2023 report last week and this showed the consistent growth of the use of machine translation across all user types (freelance translators, language service providers and corporates).  In general I don’t think there is any loss of appetite for technology and certainly AI is going to deliver more of it.

To answer these questions around how we’ll be affected and what steps we need to take to embrace it I decided that rather than share only my own opinion it would be interesting, and possibly thought provoking, to involve the very technology we’re all playing with.  The answers, whilst perhaps not surprising, do help draw some attention to what we should be doing to embrace the change.

The next two sections are the result of a “discussion” I had with ChatGPT using the GPT-4 engine that was released last week.

The effect of AI on the role of translators and interpreters in the coming years?

AI is expected to have a considerable impact on the roles of translators and interpreters in the coming years, leading to changes in their daily tasks, required skills, and job opportunities.  Here’s an analysis of how AI may affect translators and interpreters:

  1. Machine translation: As machine translation (MT) tools become more sophisticated, translators will increasingly use these tools for first-pass translations, especially for simpler, more repetitive content.  Translators will likely focus more on post-editing machine-generated translations to ensure accuracy, context appropriateness, and stylistic quality.
  2. Real-time interpreting: AI-powered applications are already capable of providing real-time interpreting services in some situations.  While these tools may not replace human interpreters for complex or critical tasks, such as legal or diplomatic interpreting, they may reduce demand for human interpreters in more casual or informal settings.
  3. Collaboration with AI tools: Translators and interpreters will need to collaborate with AI tools, including machine translation engines, voice recognition software, and terminology management systems.  This collaboration can help improve overall productivity and open up new opportunities, such as language consultancy and technology integration.
  4. Specialisation: As AI tools become more proficient at handling general translation tasks, translators and interpreters may need to specialise in specific domains, such as legal, medical, or technical translation, where their expertise can add value beyond what AI can achieve.  Specialisation can help translators and interpreters maintain a competitive edge in the job market.
  5. Focus on creative and complex translations: AI tools still struggle with translating creative works, such as literature or marketing materials, that require an understanding of nuances, context, and emotional impact. Translators will likely continue to play a significant role in translating these types of content.
  6. New skills development: To stay competitive in the evolving translation and interpreting industry, professionals will need to develop new skills related to AI and technology.  This may include learning to use AI-powered tools effectively, understanding the limitations of machine translation, and being able to adapt to new technologies and workflows.
  7. Quality assurance and control: Translators and interpreters will play a crucial role in ensuring the quality of AI-generated translations and interpretations. They will need to review and refine AI-generated content, ensuring that it meets the required quality standards and is contextually and culturally appropriate.
  8. Ethical considerations: As AI becomes more integrated into the translation and interpreting process, professionals will need to consider ethical issues, such as the potential biases in AI algorithms, the consequences of over-reliance on AI-generated translations, and the responsibility of human translators in the AI-assisted translation process.

The effect of AI on the role of a localization engineer in the coming years?

AI is also expected to have a significant impact on the role of localization engineers in the coming years, with potential effects on their daily tasks, required skills, and job opportunities.  Here’s an analysis of how AI may affect the role of localization engineers:

  1. Automation of repetitive tasks: AI-powered tools can automate many repetitive tasks in the localization process, such as file format conversions, text extraction, and quality assurance checks.  This can free up localization engineers to focus on more complex and high-value tasks, such as troubleshooting, process optimisation, and tool development.
  2. Enhanced machine translation: As machine translation improves with AI advancements, localization engineers will need to work more closely with these tools.  They may be required to integrate machine translation systems into localization workflows, train custom translation models, and perform post-editing on machine-generated translations to ensure quality and context appropriateness.
  3. Language model adaptation: AI-driven language models will increasingly play a role in the localization process.  Localization engineers may need to adapt and fine-tune these models for specific domains, industries, or clients to achieve the best results.
  4. Integration of AI tools: Localization engineers will need to be proficient in integrating AI-powered tools, such as translation memory systems, terminology management systems, and machine translation engines, into the localization process.  They may also need to develop custom solutions or workflows to maximise the benefits of AI technology.
  5. Advanced quality assurance: AI can be used to develop more sophisticated quality assurance (QA) tools that can automatically detect linguistic, formatting, or functional issues in localized content.  Localization engineers will need to understand how to use these tools effectively and address any issues they identify.
  6. Skills development: To remain competitive in the evolving localization industry, engineers will need to develop new skills related to AI and machine learning, such as programming, data analysis, and model training.  This will enable them to adapt to new technologies and contribute to their development and implementation.
  7. Focus on creative problem-solving: As AI takes over routine tasks, localization engineers may be required to focus more on creative problem-solving and process improvement.  They may need to work more closely with other team members, such as translators, project managers, and developers, to ensure efficient and high-quality localization outcomes.
  8. Shift in job opportunities: With the increasing role of AI in localization, there may be a shift in job opportunities, with a higher demand for localization engineers with AI expertise and a reduced demand for those focused solely on traditional localization tasks.

Conclusion

The elephant in room that we must all talk about is how AI is likely to transform the roles of translators and interpreters and localization workflows in the coming years.  We must think about how AI presents opportunities for professionals in the translation, interpreting, and localization industry and not just dwell on the negative, or ignore it.  I think if we have not thought long and hard about the future before and have just ignored what has been going on around us, then the rapid increase in this type of technology today is surely going to force this kind of thinking.

The technology is definitely attractive, but I’m sure we are going to see a lot of hype that will spark debate in the usual translation forums and conferences as companies seek to be seen as the ones taking advantage of AI technology to encourage investors and technology buyers to spend their money on them.  I think the reality is that AI and humans possess different qualities and abilities, and any assumption that AI can replace humans completely fails to consider these differences.  AI-based machines are fast and more efficient in some aspects, but they still lack the emotional intelligence and human connection that is essential in many situations.  In the localization industry, language translation is a significant task that requires not just linguistic knowledge but also cultural understanding and context.  AI can analyse and process data, but it is limited to the data it receives and may struggle with idiomatic expressions or cultural nuances.  This suggests that, while AI can be a valuable tool to assist human translators, it is unlikely to replace them entirely in the near future… although I do think, as noted in this article, that we will all have to change.

I also think some of the hype will be accurate too, particularly around processes that are more data driven.  Even within the translation process we can already see how this type of technology can provide super powers to anyone puzzling over the more technical aspects of the localization process.  Right now it’s a tool we can all use for our own benefit, but it won’t be long before our work requires more collaboration with AI tools because this will be the skillsets our employers and customers provide and demand.

Over the years I’ve read many articles and listened to many presentations from professionals in the field arguing successfully in my opinion that the translation profession deserves to be recognised and respected for the specialised knowledge, skills, and dedication to quality that its practitioners bring to their work.  Going forward, the rapid advancements in AI present numerous opportunities for professionals in the translation, interpreting, and localization industries.  Embracing these opportunities, rather than dwelling on the negative aspects of change, is vital for remaining competitive and relevant in the evolving landscape.  By specialising in specific domains, focusing on creative and complex translations, developing new skills related to AI and technology, and collaborating with AI tools, professionals can optimise their productivity and open up new avenues of work.  Ensuring quality control and addressing ethical considerations further emphasise the importance of human expertise in the AI-assisted translation process.  Ultimately, focusing on the opportunities presented by these changes enables professionals to adapt and thrive in the face of the industry transformation we are embarking upon.

Do you agree or not?

2 thoughts on “The elephant in my room…

  1. A very thought-provoking article, Paul. Certainly a lot to think about, both as a translator and as an ordinary citizen. Job loss is certainly a biggie in my mind, and world economies need to be proactive to reconfigure productivity sectors with this in mind, offering retraining programs and other incentives to help people keep up with change. It’s a huge area that should not be left to chance.

    1. Couldn’t agree more. Something I didn’t mention as it was off-topic a bit but relevant in the context of your comment is that while AI and automation will impact the labour market, they will also drive the creation of new opportunities in areas that complement or are less susceptible to automation and some even to ensure we are able to sustain the use of the very AI technology we’re putting in place and to support our changing needs as a human race. I’m sure we don’t even know the full extent of this yet and I’m sure it’ll change the way we live in many ways.

Leave a Reply