The translation profession vs the translation industry: a growing schism

Apart from the seismic intrusion of LLMsA large language model (LLM) is a type of artificial intelligence (AI) algorithm that uses deep learning techniques and massively large data sets to understand, summarize, generate and predict new content. More and AI into translation, there is a growing schism between the profession and the industry. At conferences, it rapidly shines through whether an event is an “industry event” or a “translators event”. I dropped in online for some sessions of #2023TEF – (Translating Europe Forum). The event is very polished – one keynote felt like I’d accidentally walked into a TED Talk (they’re not my thing). The other attracted a lot of comments from the professionals in attendance questioning its appropriateness.

Industry events focus on what tech can do (and sell it that it can do everything) whereas the professional side prefers to rationalise what is possible. It also openly questions the genuine time-saving of the end-to-end use of technology.

Industry practices have been at odds with the profession in recent years, and their paths are certainly ever more divergent. As a brief sample of some of the issues percolating between the two sides, here is a brief list. This is one reason for the increasing sentiment of “Love the profession, hate the industry”.

Where industry and profession collide/diverge

1. Translators being “reduced” to post-editors of machine translation (PEMTPost-edited machine translation (PEMT) involves human editing of machine translation. More/MPTE). Previously translators were dominant in the translation phase of projects. Many now fight for scraps in post-editing for pennies a word. Re-translation, where the MT output is not fit for purpose, is also charged at this dumping rate.
2. Weaponisation of fuzzy matches in CAT tools to drive down rates. Many CAT tools use bands of fuzzy matches of translation units. They overpenalise/discount fuzzy matches or apply discounts to fuzzy matches e.g. in the 50-74% band. Many TUs only useful as a starting point improve the margins of the industry to the detriment of the translator. Even 97% fuzzy matches, e.g. a negation, may require considerable revision. It is not just a case of simply negating the verb. Even 100% matches may still require considerable work – due to their contextual relevance.
3. Exploiting dumping rates for human translation in certain language combinations. In some cases, rather than take an MT and PEMTPost-edited machine translation (PEMT) involves human editing of machine translation. More approach, industry outsources to countries with dumping rates for translation. Suddenly German to English translations are handled by someone with neither source/target language native level mastery. And agencies then only involve someone to edit/revise with the necessary mastery.
4. Fastest finger first – the rise of “click factories”. In many instances there is no real bidding process or supply/demand pricing in a language combination. Translation allocation is strictly on a “fastest finger first” basis. The first agency to click that they can take a job gets it – there is no discussion about rates etc. Consequently ther new link in the chain simply clicks on a job the instant it appears. Some disreputable agencies now have dedicated “clickers” to ensure that they are the quickest to click.
5. Machine in the LoopMachine in the loop is an approach to human/machine translation. Under Machine in the Loop, a human expert (i.e. a translator) makes use of computer-based tools to support them in the translation process. This approach contrasts “human in the loop” (HITL), which only has subordinate human involvement. HITL also does not state whether the human involved is necessary an expert. More vs Human in the LoopHuman in the loop is an approach advocated by the translation industry, whereby machine translation is used and the role of human translators is secondary – e.g. doing some kind of post-editing. It has become particularly prevalent with the rise of LLMs. My issues with HITL is that it reduces the role of a human (with no mention of their More. Professional translators strongly oppose the “Human in the Loop” (HITLHuman in the loop is an approach advocated by the translation industry, whereby machine translation is used and the role of human translators is secondary – e.g. doing some kind of post-editing. It has become particularly prevalent with the rise of LLMs. My issues with HITL is that it reduces the role of a human (with no mention of their More) approach the industry embraces. HITLHuman in the loop is an approach advocated by the translation industry, whereby machine translation is used and the role of human translators is secondary – e.g. doing some kind of post-editing. It has become particularly prevalent with the rise of LLMs. My issues with HITL is that it reduces the role of a human (with no mention of their More relegates human translators to post-editors of machine translation – making their involvement a mere footnote or afterthought. Professionals if anything advocate a “Machine in the Loop” (MITLMachine in the loop is an approach to human/machine translation. Under Machine in the Loop, a human expert (i.e. a translator) makes use of computer-based tools to support them in the translation process. This approach contrasts “human in the loop” (HITL), which only has subordinate human involvement. HITL also does not state whether the human involved is necessary an expert. More) approach. Under MITLMachine in the loop is an approach to human/machine translation. Under Machine in the Loop, a human expert (i.e. a translator) makes use of computer-based tools to support them in the translation process. This approach contrasts “human in the loop” (HITL), which only has subordinate human involvement. HITL also does not state whether the human involved is necessary an expert. More, human language professional retains their role, but with recourse to technology in the process.
6. Collaboration with AI: the industry ihas been quick to embrace the use of AI. It has made sweeping statements such as the death of translation as a profession as we know it. It also propagates the “truth” that “if you don’t use AI, you will lose out to a professional who does”. AI is rapidly becoming kryptonite for professionals. It is making an in-road in their earnings, with a dwindling amount of work for translators.

I remain firmly on the side of human expertise and advocate a quality over quantity approach, as well as ensuring proper remuneration for translators. So yes, I love the profession, but hate the industry.