Free Consultation
May 14, 2020

Rage Against the Machine: Human Translators vs. The Robots

A mentor of mine has been encouraging me to think about the threat that machine translation (MT) poses to human translators and translation agencies alike for a long time. 
Every time, I respond that despite machine translation fulfilling a larger portion of the demand for language services every day, the demand for translation services is, overall, growing at a steady pace. However, it is worth going beyond this simplistic statement and reviewing the current state of machine translation.

Where is MT at today?

Machine translation has improved dramatically over the last decade. From the early days of Google Translate to the emergence of a wide variety of tools, a lot of resources have been invested into R&D to make it easier to understand foreign languages and, more and more, generate content into a target language. Broadly, there are two categories of tools:
  • Online free automated translation: Google Translate is the leader here and the most commonly referenced. In recent years, other services such as Linguee or DeepL have been developed to challenge Google's dominance. Those tools work best on very structured content, such as simple and common sentences but also (surprisingly) on agreements, patents, and technical content. The main drawback to using them is the leak of proprietary information (since your content will then become part of Google's reference material). The output is also still far from perfect so to this day online free translations are mostly suitable for comprehension purposes.


  • Proprietary automated translation: Some large agencies realized that they had built considerable translation memories (TM) for their clients over the years, and that the TMs could be leveraged to generate automated translations and potentially reduce the cost of translation for clients. One example is Star Group’s STAR MT. Clients can receive automated translations based solely on their source material, knowing that their information is not processed via public sites. Review by human translators is however necessary to “fix issues”, a task many translators are still reluctant to perform. 

Are human translators obsolete?

Every business sector is affected by automation. According to former POTUS candidate Andrew Yang, Yes, Robots Are Stealing Your Job”. While we have accepted that some professions already faded into irrelevance (lift operators, cashiers, airport counter staff…), humans will continue to be relevant for a number of important reasons:
The first reason is about risks and liabilities: while reducing the cost of translation can definitely be meaningful for many companies, there can be hidden costs. Errors introduced during the translation process can cost companies a lot of money. If the translation was done for free by a robot, one can only blame themselves for using an automated system. 
Emotional intelligence is also an important factor: for the very same reason that content is one of the most desirable commodity, the translation of content will continue to be valuable too. Machines are not (yet) able to replicate the traits and emotions that make us who we are: showing empathy and kindness, understanding other people's needs based on one's own experience and background, being creative and adapting to complex situations involving other humans. Those are all skills at which translators excel, as they always operate between two cultures (or more).
Finally, we need to remember why companies choose to translate content into another language in the first place. Businesses want to show that they care so much about their foreign users, that they are ready to produce impeccable content in their native language just to communicate with them. Taking the risk of sharing content that is ridden with errors and inaccuracies would have the opposite effect. 

Machines and Humans address different needs

There is no doubt that machine translation will continue to improve and - probably - replace a portion of the work translators used to do on translations aimed at basic comprehension. I would argue that this is fine. Companies need to use their resources wisely after all.
On the other hand, the same companies will likely want to communicate the message of their brand in an impeccable, culturally sensitive way. This is where translators will continue to add tremendous value by diligently translating every word, and conveying human traits such as empathy and dedication.

Get updates on using language translation to globally scale blockchain initiatives:


March 2020

Timeline for translation projects

Mar 10, 2020

December 2019

Translation: freelancing or in-house?

Dec 2, 2019

April 2019

Project Management at Kotoba

Apr 1, 2019

February 2019

How to Apply as a Translator?

Feb 1, 2019