Why we need XITL in addition to MITL and HITL

I’ve been pushing the case for “Experts in the Lead” (XITLEITL/XITL advocates the expert human in the lead rather than in the loop.) a lot recently, as a new term for human-led translation. It implicitly defines that humans remain in the lead, while “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.) only infers it. In contrast “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) relegates human involvement to a trivial nature. A parallel that ITI seems keen to push is Francois Grosjean’s quotation that two hands and ten fingers don’t make you a pianist – just as speaking two languages doesn’t make you a translator. Your expert in the leadEITL/XITL advocates the expert human in the lead rather than in the loop. is a translation virtuoso.

Why do we need it?

Currently two terms crop up frequently in relation to human-machine translation. Both are quite weaselly for the translation profession: machine in the loop (MITL)Machine 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. and human in the loop (HITL)Human 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. Both indicate a continued role for humans in some shape or form (i.e. some task left to the human). However, neither acknowledges the need for human expertise that machines are not capable of. 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. indirectly infers that a human remains in the lead in that a machine is only in the loop. In contrast, 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 directly states mere (non-expert) human involvement. Check out my assessment of the state of play at the start of 2024.

Both these terms suit the translation industry well, but do little to assuage the concerns of the profession. This is why we need a third option. The necessity of the Expert in the Lead (XITL)EITL/XITL advocates the expert human in the lead rather than in the loop. approach is what the language profession needs to emphasise. It isn’t about being a Luddite and rejecting technology. Many experts in the lead have used CAT tools for decades. Both Trados and MemoQ celebrate round birthdays in 2024, turning 40 and 20 respectively.

What else do we need it for?

In advocating an expert-led approach, we should also promote technological agnosticism. Human experts in the lead should be free to decide how and which technologies they use. A cuvee depends on the specific blend of grapes – expertise also needs to find the perfect balance.

XITLEITL/XITL advocates the expert human in the lead rather than in the loop. approaches won’t sit well with LSPs exploiting 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. and 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 for larger margins. It will however justify better rates for human experts. Commoditisation of translation into characters, lines and words is often part of the reason why customers look towards 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. Past translation quality may have been a driver for the industry to look to at new ways to earn. Mediocre results at premium rates also create a market for customers looking for “good enough” results. This is where real expertise needs to come in. To achieve this translators need to also make sure they convince their customer’s decision-makers. In another of my recent blog posts, I discussed Chris Durban’s clarion call for translators “to visit Clientland”.

When do we need it? (Now!)

Expertise and experts need visibility and being heard above the industrial noise. Think of it like ensuring that a building site doesn’t operate around the clock in a residential area.

We’re also not talking about job title inflation (à la “freelance translators” becoming “professional translators”). The experts in the lead revolution needs to see people proving their expertise. A recent CAMELS interview with Deborah Fry highlighted a need for specialist subject-based training by subject matter experts. This is the way forward rather than an “opiate for the masses” type approach.

Many large professional events have tracks on the work/life balance side of translation. This is all well and good, given the cognitive demands of translation, but does not assist expertise building. I struggle to attend conferences, where the added value in terms of subject matter expertise is not obvious.

I have to convince line managers why I need to attend events. The rationale is not merely financial. Time out of the office plays a big role if the “red line” for enhancing my expertise isn’t apparent. As does remuneration for attending weekend events. There is a time and a place for popular subjects covered at conferences. Such events contribute to professional development, as does networking, but they can fall short regarding subject matter expertise.

How do we go about it?

From my holiday reading, Cal Newport’s “Deep Work” from 2016 has certainly struck a chord. The Expert in the LeadEITL/XITL advocates the expert human in the lead rather than in the loop. needs “deep” or focused work rather than shallow work (i.e. like PEMTPost-edited machine translation (PEMT) involves human editing of machine translation. for peanuts). The book advocates concentrated work, away from distractions (social media, e-mail, instant messages). Projects with substance help in this regard too – rather than fighting over scraps. Refining processes – like establishing better briefs can also help.

Similarly, we need to think about the battles we fight. Picking up those MT fails and sending them viral isn’t where it is at. We need to focus on how to improve and extend our expertise. I’m quite lucky in that the expanding remit of supervision means new supervisory areas (e.g. DORA, MiCAR, ESG). And the transpositionThe act of transposing European law (e.g. Directives or Regulations) into national law. It should be noted that European Regulations are directly applicable, whereas Directives require transposition to be applicable in national law.. of CRD6 and CRR3 into Austrian law is a fresh seam of translation content at the wordfaceA play on words with the "coalface". Translators could be said to work at the wordface and might suggest "Getting back to the wordface!" to mean get back to work. The term apparently originates from Terry Pratchett's "Going Postal". A few feet beyond there was the familiar wall of letters, packed as tight as rock strata. A watchman had been.

Examine the way your area of specialism is going (and the next big things) and proactively obtain expertise. This is what “staying good” is all about. And if you aren’t good yet, devote as much time as possible to getting good, quickly.