A couple of weeks ago, I read the new AI in Translation ebook from Nimdzi. (It’s available in a number of places online; let me know if you can’t find it and I’ll be glad to email you the PDF.)
It was valuable to hear perspectives from across the industry, and Joe Lepine—yes, THAT Joe Lepine—did a diligent job defending translators. But among the Localization Illuminati, he often felt like a lone voice in the wilderness.
I’m sure the so-called industry leaders are nice people doing their best during an extraordinarily difficult transition. Still, many of their claims don’t match what I’m seeing.
We’re told that although entry-level translators are being hit hard, demand for veteran linguists, editors, transcreators, terminologists, and quality leads is growing—that “the human in the loop gets repriced upward.”
Really? Where is this thriving market for veteran linguists at substantially higher rates? Encouragement isn’t evidence.
The book also celebrates “AI-first” delivery while promising “human judgment at the center.” What does that actually mean? Who designs the process, makes the decisions, and captures the value?
Similarly, translation memory is supposedly depreciating, yet clean underlying data is essential. If that data isn’t translation memory, what is it? Who creates it, maintains it, and benefits from it?
Then there’s this: “The linguist of the present is a designer of outcomes.” It sounds impressive, but how many LSPs let linguists design anything? Many won’t even engage with our feedback, let alone invite us to architect workflows or raise our rates. Those jobs belong to engineers and internal teams.
That’s why telling translators to become “process architects” and “workflow designers” can feel like gaslighting. It describes an attractive future without explaining how we gain the authority, access, or compensation to inhabit it.
The discussion rings truer when it focuses on human judgment. Our judgment is real, hard-earned, and valuable. But how will it be integrated, recognized, and paid for?
Eventually, the book admits that translators’ livelihoods are being “reshuffled” and that we’re owed an honest conversation about what comes next. That may be its most honest statement.
The problem is: I don’t think industry leaders know what to tell us. That’s unsettling—but also clarifying. We cannot wait for LSPs, technology companies, or commentators to define our future.
There is no easy path. All the options are hard.
Some translators are finding a way through by learning new tools, strengthening client relationships, finding new markets, deepening their expertise, and taking greater control of their workflows.
No matter who we are, we must keep learning and experimenting.
That has been my message from day one. It underlies Thursday’s workshop with Joe and Anne, and everything we’re building with CotranslatorAI. But these are only beginnings. We still have much to discover about where our value lies and how to deliver it directly to clients.
Translators aren’t alone: small LSPs and mega-agencies are struggling too. Nobody has this figured out.
That’s cause for some despair—but also for action. We must question the narratives, test new models, share what works, and claim an active role in shaping what comes next.
