When we talk about the rise of AI in language services, the conversation too often gets stuck on automation. Most stakeholders, such as language service providers, scholars, and decision-makers, tend to view technology primarily as a tool to streamline existing tasks, making familiar processes faster, cheaper, or more efficient. But in my view, that’s the less compelling, and ultimately less sustainable, side of technological transformation. Automation is a path that surely offers short-term gains for companies, but one that, if it remains the only strategic focus, might lead to long-term obsolescence.
The real shift for a sustainable business future, I argue, lies elsewhere. It’s not about doing the same things better. It’s about doing entirely new things. Things that were previously unimagined and impossible are now achievable, thanks to technology’s capabilities and, in the context of languages, AI. That’s not automation. That’s innovation. Pondering about the differences between the two seems to me more important than ever.
To fully grasp the significance of this distinction for the language industry and the profession, it’s helpful to recall what Richard Susskind, renowned scholar and author on AI and the future of work, argues in his 2025 book. He draws a sharp line between automation and innovation, and insists that the future of professional services will depend not on how well we preserve old models, automating them as much as we can, but on how boldly we embrace new ones.
Automation, Susskind explains, is about using technology to replicate human work, often more efficiently, often at scale. It involves the mechanical reproduction of tasks once performed by people, in our case, by professionals. Machine translation is the textbook example: where there was once a human translator, we now have (or might have) an AI engine. The input and output remain largely the same. In the automation paradigm, the human is simply removed from the loop.
There are many reasons why individuals and industries pursue automation: to cut costs, boost margins, or improve quality and speed. This also applies to the language industry, both to translation (see Moorkens and Guerberof Arenas, 2024) and to interpreting (see Fantinuoli, 2019). But innovation is something more radical. Susskind defines it as “delivering the outcomes that clients or users have long sought, but using techniques or technology that support radically new underlying processes”. In other words, it’s not just about doing the same thing better or cheaper: it’s about reimagining what’s possible.
“To think of AI, as most people do, simply as a tool for automating today’s tasks and activities is enormously to underestimate its current and future impact.”
— Richard Susskind
Why is it paramount to reframe technological transformation in the language industry as innovation rather than mere automation?
Automation has intrinsic limitations (see for example Svanberg et al., 2024). By definition, it is bounded, both in ambition and in outcome. First, once you have automated a process, you’ve reached a plateau: your goal has been achieved, and further improvement becomes incremental at best. If your business model or competitive advantage relies solely on automation, growth is capped. The total gains are constrained by the proportion of tasks that can be automated. Once that ceiling is reached, progress stalls, and decline often follows. In simpler words, if our strategy to remain relevant is more automation, when we have automated everything we can, then our strategy will not produce any more value.
Second, automation rests on a flawed assumption: that all jobs can be reduced to a set of discrete, automatable tasks. This “bundle-of-tasks” model, popular in labor economics, however, is fundamentally incomplete. When we define jobs strictly in terms of tasks, we risk overlooking the most complex, nuanced, and interdependent aspects of human work, the parts that exist at the boundaries between tasks. It doesn’t matter how advanced AI becomes at standardized benchmarks; the messy, often underspecified nature of real-world work is precisely what makes it resistant to full automation1. In simpler words, sometimes we expect simply too much from our automation efforts.
These reflections on automation should also give pause to the language services industry, which, I would argue (perhaps wrongly), remains largely focused on automation. In a time when professionals and organizations alike are grappling with uncertainty about their role in the future, both in terms of identity and of economy, doubling down on automation may surely offer short-term efficiency, or even profitability, but in the long term, it risks leading to obsolescence.
By contrast, innovation – doing things differently, or doing things that were never possible before – opens up entirely new pathways for value creation. Innovation knows no ceiling. There is literally no finish line for how far we can innovate. Unlike automation, which eventually reaches its limits, innovation holds the potential for continuous reinvention of markets, services, and user experiences. It’s an open-ended process of creating new value, not just optimizing the old. Once you have innovated, you can innovate again, and again.
Of course, automation and innovation always coexist; automation can be a component of innovation, and vice versa. The distinction isn’t always clear-cut. But if we want to avoid the dead ends of technological automation, giving short terms business advantages, but leading to decrease of values after each cycle of automation, we must actively shift the frame. Bear with me, and let’s explore what this reframing might look like in the language sector, with examples that move beyond efficiency to reimagine the very purpose of our work and strategy.
One compelling way is to shift the conversation from replacing human translators and interpreters to transforming the very experience of multilingual communication. As we have introduced above, in translation and interpreting, innovation means moving beyond replicating human tasks toward creating entirely new forms of value. For instance, instead of offering One-Size-Fits-All solutions, we could envision Adaptive, On-Demand Translation Services tailored to the user’s profile, context, and purpose, something that traditional workflows could never deliver. Two examples.
Imagine you’re reading a foreign article online: only this time, the translation hasn’t been predetermined by the publisher as a one-size-fits-all version for every reader. Instead, imagine something that we could name fluid translation: the system is adapting the text in terms of language, tone, and even examples to suit your background, your profession, your level of expertise. Legalese becomes digestible for a layperson. Technical jargon becomes relatable. Images are used for a person that prefers visualisation supports rather then verbose messages. Instead of having one single version, translation becomes a personalized experience, dynamic and real-time, generated ad hoc for the specific reader.
Or picture a scenario in simultaneous interpreting. Instead of providing the same version of a speech to an entire audience, the system offers a personalized interpreting experience: a medical professional hears terminology aligned with their field; a policymaker hears the speech with regulatory implications emphasized; a person with cognitive disabilities hears a simplified version. This isn’t fantasy. It’s what becomes possible when we stop focusing on simply replicating human processes and start inventing entirely new ones, when we shift from an automation mindset to one of innovation.
Innovation Means Rewriting the Rules
True innovation in our field doesn’t ask: How can we make translation faster or cheaper? It asks: What if translation were no longer a fixed product, but a fluid service? What if users didn’t just receive a translation, but interacted with it, influenced it, customized it?
That’s not automation. It’s a fundamental redefinition of how language services are conceived, delivered, and experienced. And this shift holds significant value, both societal and economic, not only in the short term, like automation, but in an indefinite long term perspective.
The Role of Experts in an Age of Innovation
As language service providers, we have a choice. We can anchor ourselves to the automation debate: fighting to preserve old workflows, to improve them, or to fear replacement. Or we can lead the conversation around innovation. Because in this new paradigm, human expertise doesn’t vanish. It evolves. Innovation invites us to imagine a future where communication is not only faster, but smarter. Not only multilingual, but meaningful: tailored, contextual, alive.
The future of translation is not about replacing the human at all costs. It’s about rethinking what it means to understand each other in the first place. And to do that — creating new value, including economic value — we must first understand what innovation truly means.
Bibliography
Beraja M and Zorzi N. “Innefficient Automation“. 2023. MIT.
Fantinuoli C. “The technological turn in interpreting: the challenges that lie ahead“. 2019. Proceedings of the conference Übersetzen und Dolmetschen 4.0. – Neue Wege im digitalen Zeitalter, pp. 334-354.
Moorkens J. and Guerberof Arenas A. Artificial intelligence, automation and the language industry, 2024. In: Handbook of the Language Industry. De Gruyter, pp. 71–98. https://doi.org/10.1515/9783110716047-005
OECD. The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis. 2016, OECD Social, Employment and Migration Working Papers. Organisation for Economic Co-Operation and Development (OECD). https://doi.org/10.1787/5jlz9h56dvq7-en
Susskind R. How To Think About AI: A Guide For The Perplexed. 2025. Oxford University Press.
Svanberg M., Li W., Fleming M. Goehring B., Thomson N. Beyond AI Exposure: Which Tasks are Cost-Effective to Automate with Computer Vision? 2024. MIT.
- A third issue with automation is obviously the well-known problem of job displacement, for which a good reference is Baraja e Zorzi (2023), or in the field of translation and interpreter the work of Giustini. ↩︎
A very insightful and realistic article about our profession, very good food for thought. thank you
I follow your forward thinking pieces with a mix of awe, excitement, and passion.
Bonne continuation
Thank you!
Learned to shift my way of thinking about this topic!