Category: For Practitioners of Translation & Interpreting

Information for translation practitioners, a breed threatened by the adoption of AI by many of their providers of work.

Two Obvious Truths + One Bonus Truth

Here are the two obvious truths:

Truth 1: The need for human translators has not been obviated by MT, with or without AI.

Truth 2: It there are 100 translators reading this, fewer than 20 or maybe fewer than 10 of them will still be able to make a living by translating after about mid-2027.

Translators incessantly harp on Truth 1, and I agree with Truth 1, but most carefully avoid even thinking about Truth 2 and what it might mean for them.

The trick is to be among the few survivors. But don’t worry about how to do that, because there’s a third truth.

Truth 3: Almost all the translators who will survive until after about mid-2027 have already done what they need to do to give them the tools to survive.

And the tools I’m talking about are not CAT tools or AI software. They’re tools acquired by making early life and career decisions. That includes doing things to acquire field-specific expertise that is convincing to clients having that expertise without the need to Google, and it also includes where to live, yes, even in the age of global connectivity from anywhere.

Since it’s too late for most people already translating to start, the task that faces working translators who are facing AI—and that’s almost all translators—is to figure out whether they already have the tools for survival and are in a situation that enables them to do the necessary to survive AI. Most do not have the tools and are not in favorable situations.

Once translators can figure that out, they can make decisions about what to do, and that includes getting out of translation or using their hard-earned language ability for something else.

Waiting to be swallowed up by the tsunAmI is not a good strategy, unless you are already financially independent and have a comfortable lifeboat at your disposal.

Dial me Yesteryear

The other day I saw a comical meme depicting a young person being puzzled as to how to use a dial telephone. If we are not there already, the day will come when almost nobody will remember ever “dialing” someone’s phone number.

I am beginning to think that there will be a day when there will be few people nominally “in translation” who will remember the days when earning a living by translating was possible. By that time in the not-too-distant future, most people in need of translation will have learned as part of their education that machine translation is available (no need to invoke the buzzword AI, since “translacide by MT” was happening way before the Great 2020s AI Hype), and that will cause a drastic deprecation of the efforts necessary to learn a foreign language.

People who persist in learning foreign languages will have been told—the universities having finally agreed to recognize reality—that there are few careers open to people who want to translate, but that they might be able to make some money post-editing machine translation output if they have no other employment paths available.

The above-noted day will surely come, but when?

Although most commercial translators working in fields with traditionally high demand will cease to receive much translation work in perhaps two to three years (probably around mid-2027 for Japanese-to-English translators), they will still remember the good old days. Some of them will move away from translation and language-related careers entirely and, when you take into account the newcomers that will have been groomed to be post-editors coming into the labor market, I would think that even the memory of actually doing translation will be pretty much gone in about 20 years.

Perhaps more than in the past, newcomers studying Japanese will have hopes of doing translation in creative or entertainment-related fields, but literature was never easy as an income-earning profession (even for original authors), and games and subtitling—highly competitive even before the glut of translators resulting from MT adoption—will not be attractive from the standpoint of earnings, however popular and exciting those fields may be.

It’s coming. In fact, it is here for some people.

Non-collaborative Collaboration

William Lise

“Collaboration” is a usually inappropriate and arguably deceptive term skillfully promoted by translation brokers to freelance translators, many of whom have adopted the term into their active vocabulary as if it reflects reality; it usually does not.

With rare exceptions, when a freelance translator says that they are collaborating with an agency, there is no more collaboration going on than when a farmer sells lettuce to a grocery shop that resells the lettuce.

The functions, if any, performed the translation broker are usually at best analogous to the grocery shop washing off the lettuce before selling it to their customers. They don’t know much about the lettuce they are purchasing for resale, and that is about the level of involvement that the translation brokers that sell most of the translation that is purchased are capable of—or desirous of—having in the translation process.

Even the lettuce-washing task (post-editing of machine-translation output for many translation brokers these days) is often given to non-employees outside the shop, some of whom might themselves be capable of farming, but usually not selling, lettuce.

Recent advances in technology have enabled translation brokers to produce their own pesticide-sprinkled lettuce, eliminating even their need to purchase lettuce, confident that there will be non-employee, out-of-work lettuce farmers standing at the ready to wash the produce off for them. That confidence is turning out to be warranted.

This is the nature of the translation-brokering business. It will only get worse for the lettuce farmers, some of whom might be able to set up and sell at roadside stands, but there is not much time left to do that, and many are loathe to even think of it.

Breaking up and Honesty

William Lise

As freelance translation orders from agencies wind down (remember, please, that some Titanic passengers were doing just fine until the end), I am surprised translation organizations have not made efforts to counsel their members on what else can reasonably be done to survive.

While it is clearly impossible for most translators even to buy time by getting direct clients that don’t use AI, there are other things that can be done, but they’re not being talked about, unless I’m missing something, and I don’t think I am.

I am not referring to offering a suite of services related to AI that are being demanded by agencies, namely post-editing and other peripheral services that are not themselves translation, these services sometimes being talked about or implied as a means of “adapting” to AI.

And I am not referring to freelance translators using AI themselves, because they would need to find clients that don’t themselves use AI, and that is impossible for most translators.

Nor do I refer to going into fields such as entertainment translation that, while interesting, attractive and popular, represent only tiny amounts of demand compared with the mainstream translation fields that have supported most freelancers in the past. In addition, such popular fields are highly competitive and don’t offer attractice income possibilities.

Alhough suggestions of the above-noted strategies provide convenient and anodyne presentation topics, they don’t address the real problem directly and miss the point of what is happening.

I refer rather to in-house jobs that require knowledge and skills other than translation and that either require language ability or are made easier to acquire if you are bilingual.

That said, many of those positions are not going to be available to people well into their careers or to people who can’t pick up and just live somewhere else, because—surprise, surprise—many “real” jobs require you to go into work, and a good portion of the freelance translator population has for decades enjoyed the ability to work at a distance from—and out of sight of—the people who pay them.

Still, I think that the translation organizations are missing out on a moral opportunity to give advice on what can be done after it’s all over, and the “all over” is  clearly happening, not-so-slowly, and surely. The same can probably be said of universities. Both translation organizations and universities appear to be opting for the appearance of relevance, even if not justified, over substantial usefulness to working translators and prospective translators, even if that usefulness boils down to telling them where else they might go. Organizations such as Japan Association of Translators and American Translators Association have opted for dishonesty by omission.

The translator-agency relationship is breaking up, which we know is hard to do, and honesty is such a lonely word, as Neil Sedaka and Billy Joel put it.