Let’s consider the biblical parable of new wine and old wineskins in the context of translation today.
Generative AI provides us with new wine (an entirely new paradigm). Yet the industry mainstream is trying to fit this new wine into old wineskins (outdated ways of working).
The result: mediocre translations that serve only one market, the cheap one.
Here are three fundamental errors being made by the mainstream narrative.
1. Translators aren’t translators anymore.
They call us post editors, quality checkers, managers of cultural nuance (whatever that is!), or anything else they can come up with that doesn’t address our role properly.
This erroneous belief pulls the carpet out from under our role.
I saw a post on LinkedIn this week saying translators in Denmark won’t be recognized as translators when editing machine-generated output.
This reflects a misunderstanding of what the new paradigm truly offers us.
2. Translation is a linear process.
If translators just check machine output, then maybe we’re post-editors.
But why do we cede first-mover advantage and let the machine go first?
When the algorithm makes the early decisions in the translation process, then the translator can only spend the rest of the job adapting and reacting to that output, often fixing the same problems over and over again.
That’s machine-translation post editing.
And whether with AI or not, thinking you can just run a document through an algorithm in one go leads to mediocre work, even if you tag on a post editor.
We’ve got to stop thinking that quality translation can be assembled in steps.
Creating excellent work demands a holistic approach. Today, this is done when the translator and AI collaborate iteratively.
The Generative AI Iterative Translation (GAIT) Workflow is a dynamic process. With GAIT, the translator crafts the initial gold-standard translation and then iteratively develops an anchor prompt over the course of the project, continually refining it alongside the AI to achieve superior results.
There’s nothing easy or automated about this effort.
Mastering it involves developing a new skillset.
But it works.
It’s faster than human-only translation.
And it’s the clear path to the highest quality translations in the industry.
3. We are on a relentless ride down to a world of free translation.
Oh, yeah? I have news for you… It’s already free; it’s been free for a long time. It’s just mediocre.
We have to stop believing that there’s only one valid narrative of the future, that translator roles can only diminish over time.
This perspective misses the hidden demand for excellence, where top-quality translations shine.
Yes, as translators learn to take command of the AI and deliver better work faster, rates will (continue to) drop even as quality rises.
But this is nothing to be afraid of.
This will drive the supply and demand curves in our favor, opening up new markets for high-quality translation that were never available before.
Who’s going to compete with that?
Hey, if the Localization Illuminati want to keep pursuing their race to the bottom, with plain-Jane translations and lower prices that compete with free…
I’m sure it will work for some of them; there’s a market for it.
But as translators, as professionals who love our work and want to return to doing everything we enjoy about being translators, that is not our future.
If you’re ready to break away from these old ways of thinking and embrace the excellence of the future, let’s start writing a new narrative together.
Keep pushing translation boundaries!