Replacement of Professional Translators by AI-using Translation Brokers: Some non-intuitive impact on what strategies are useful.

Seeing human professionals being rapidly replaced by technology in the translation process, translators are responding in various ways, but many of those ways involve using technology or being used by technology users.

There are the adaptists and the adoptists, the former essentially surrendering to mind-numbing post-editing tasks that pay ridiculously low, and the latter calling for pro-active use of AI by freelancers to translate. There is actually another group—it’s made up of optimists who believe that clients who value human translation will remain, which is arguably not a total falsehood, but they seem never to discuss what type of enlightened client they envision, nor do they think about how very few freelancers will be supportable by that quickly disappearing client demographic.

None of these groups seems to want to think about just who will purchase their translations after the only client group they have—it’s made up of almost exclusively of translation-brokering agencies—has totally disappeared, a situation which is well underway to becoming a reality.

It would be useful for the members of these groups of optimists of various types to do some thinking about how to develop clients that have not moved away from humans yet and might not move away for at least some time. Some time can still be bought, but not by simply doing the no-risk things that have become the norm in freelancer behavior.

At least here in Japan, but surely to a great extent elsewhere as well, developing those clients boils down to doing things that don’t involve much technology—in short, it amounts to having a life outside of your computer screen and your online presence as a translator.

Face-to-face encounters in Japan are much more likely to turn into orders than developing a killer profile on LinkedIn or having a killer website, neither of which has much sales value in Japan.

Sadly, getting a life is not an option for many freelancers, precisely because they have chosen a lifestyle—it would be wrong to characterize it as a business model—that provides a working environment that makes it difficult to get a life outside of waiting in front of your computer for work to arrive.

Rather than complaining and engaging in commiseration about post-editing rates or advocating surrender terms, the energies of freelancers and the groups that purport to represent their interests would be better spent thinking about actionable survival strategies that can be adopted.

The way things have developed, those strategies don’t need and are not helped much by using technology to develop a personal brand online or doing a whole variety of things proposed by people trying to avoid getting a life in real life or trying to pretend things will be just fine. They won’t for most freelancers.

Nonsense on Steroids

From Google, upon being asked whether freelance translation is ending.

Freelance translation is not ending, but it is undergoing a massive transformation. While generative AI and machine translation (MT) have eliminated low-end, commodity translation work and depressed rates across many agencies, experts and experienced linguists remain in high demand. The market now requires professionals to shift from translating raw text to providing specialized, high-value services.

Huh? I guess I should not be surprised to see this self-serving Google AI overview when I asked it the rhetorical question of whether freelance translation is ending. Google is heavily into promoting AI, and it is best (for them) not to characterize AI as anything other than an uplifting technology that is reconstructing our world in a positive way.

What generative AI and machine translation have eliminated is definitely not limited to low-end, commodity translation work. Rather, it’s translation of subject matter that requires a human to understand it and be able to write for readers familiar with same.

Does this Google collection of software commands posing as intelligence “think” that patent translation is low-end, commodity translation? How about legal, business/finance, medical, and pharmaceutical? How about a 100-page document describing the operation of an MRI system?

These are the mainstream fields that the great majority of freelance JA-EN translators have made a good living from for decades. We are now told that they represent low-end, commodity translation. Well, yes, after the translation-brokering agencies reduce their evaluation of translation in those fields to peanuts because their AI system, followed by repair work done by desperate former translators will work for them and their clients, and it is indeed working.

“Experts and experienced linguists” are in high demand? Well, perhaps, but only because most are trapped in a captive labor pool that can go nowhere else to obtain gainful work, including competing with the AI translation broligarchs for direct clients, and are willing to work for a fraction of what they could have earned before having their lunch eaten by AI-using translation brokers. And those newly required services are claimed to be high-value, but they are being paid insultingly low, which reflects their true value as seen by the purchasers.

Being called a linguist is not going to pay any bills for mid-career translators who have lots of mid-career life expenses and had expected to be able to live without a financial threat from technology. The story doesn’t end, however, with Google’s off-the-mark take on the effect of AI on freelance translating.

It’s bad enough having places like Google feed you this nonsense. Organizations purporting to represent the interests of translation practitioners are falling in line, chanting the “new role” mantra and acting like people should just roll with the punch and take on those exciting new roles. That is simply irresponsible and contrary to their professed mission statements, but I guess these people need to pretend they and the organizations they run are still relevant, and to hell with the practitioner members who are seeing their relevance rapidly fading away.

Discount for Handwritten Translation?

One day sometime in the early 1990s, a new manufacturer client ordered the translation of a short description of a manufacturing process. I quoted my usual rate for such translations and told him I could have the translation back within about three days.

The client came back and asked me if I could do the translation for a lower rate if he could accept a handwritten translation. I needed to break the news to him that a handwritten translation would not only require a bit more time but also a higher fee.

At the time, the keyboard was still sensed by some Japanese people as being an impediment to writing. I believe most people have now overcome this view.