Business Survival as Translator:Aspirational Delusions and Myths about AI Translation
by William Lise (June 26, 2024)
There is much talk of AI among translators recently, and also many aspirational misunderstandings of the real world of the translation business. It is time for translators to take stock of what is happening around them, rather than to aspire to something that is highly unlikely to happen.
When clients realize how bad AI is, they will come back.
What clients? If this aspirational delusion is talking about translation agencies, it is simply wrong. They will just demand better post-editing, most likely without much increase in rates, because they have a sufficient supply of docile labor that cannot go elsewhere.
And if the clients are direct clients (translation consumers) of agencies, it is highly unlikely that they will "come back," other than to their agency vendors with demands to fix quality problems. Almost no direct clients are going to look anywhere but to their agency or another agency. Why? Because that's what they've been doing all along and they don't even know about freelance translators.
AI translation is not good enough for things that matter.
This is nonsense and is a serious disconnect with the reality that poor translations have been used for things that matter for decades.
Professional translators serve as bridges between cultures.
True for a tiny portion of translators. The overwhelming portion of translation that is purchased, however, has nothing to do with culture, that overwhelming portion consisting of translations for domains such as the following:
- Financial/business
- Technical/industrial
- Patents
- Pharmaceuticals/medical
- Legal
A patent about a photolithography process for use in semiconductor manufacturing has nothing to do with culture, as is the case for almost all translations in the above domains. The "bridge between cultures" claim sounds impressive, but it is descriptive of almost none of the translations that are purchased.
Anime, manga, games and entertainment media are high-profile domains and, along with marketing content, have cultural aspects, but these domains have only a tiny demand and keep very few translators busy, compared to the above-noted and other, commercially more-important demand sectors.
You can survive as a translator if you adopt AI.
Not if your only clients are agencies that are using AI themselves. And, ultimately, not if you have direct clients that are adopting AI for their translations. It is unfortunate that the simple adoption of AI to survive is being promoted by translators and translation organizations that should know better.
Lives will be lost because of bad AI translation.
This is pure nonsense. In almost a half-century of translating, I have not heard of a single life that was lost because of a bad translation.Litigation will result from mistranslations by AI.
First, see above. Second, remember that the use of AI removes the last vestiges of personal responsibility and accountability on the part of a specific human. The threat sounds good, but it is a demonstration of cognitive dissonance.
Agencies will go out of business by adopting AI.
It appears that agencies have adopted AI full-steam ahead. I haven't yet heard of one failing because of AI, but agencies have commonly failed because of other reasons.
AI is fully capable of replacing professional human translators.
This one requires a bit of nuance. From a practical standpoint, this is true, but it requires the condition of a sufficient supply of capable post-editors. That condition will continue to be met.
AI will fail because it doesn't understand anything.
The part about AI not understanding anything is true, but that itself will not cause AI to fail, because the underlying condition in the AI translation business model is the continued adequate supply of professionals who will do post-editing. Every indication is that the condition is being met and will continue to be met.
AI translation companies will not be able to operate because they will fail to secure the services of qualified professionals to do post-editing.
Because translators in mid-career often have financial obligations and are too far along in their work life to easily reinvent themselves, but not far enough along to quit working entirely, the supply of docile post-editors is not likely to dry up anytime soon.
As long as your agency clients are not using AI, you will be safe.
True, until your agency clients realize that they are being beaten in the market by agencies that use AI, thereby leading them also to use AI.
As long as you have direct clients, you will be safe.
Well, you might be safer for a longer time than if you were agency-dependent, but direct clients will also adopt AI, at which point the hopes of a translation career for most people will fade away.
Rates for post-editing will increase to the extent that you do not have to look for another job.
You need to think what would cause an increase the post-editing rates, and that comes down to labor supply and the desperation level and resources to survive possessed by the post-editor labor market. The realities surrounding those rate-determining factors strongly suggest that the rates need not be increased significantly, because the desperation level is there and many translators do not have other options.