Artificial Intelligence Learns From its Mistakes

It is well known that we all learn from our mistakes. All creatures from the animal kingdom answer a causality principle which forces them, by memory, reason or instinct, not to reproduce an error if they have already been subjected to a physical or psychological reprimand. Since this type of perception is specific to beings endowed with feelings and consciousness, it seems impossible to apply it to computers. But what about artificial intelligence ?


Nowadays, AI is not only limited to the recognition and processing of programs made to achieve a specific goal, one action at a time. In 1951, at the beginning of AI, a student from the University of Manchester created a machine capable of beating you hands down at the game of chess. This is an example among many others of primitive artificial intelligence, outstanding at the time, but quickly outdated by technical and scientific needs.

Through the years, scientists have discovered the need to revolutionize AI and decided to use the most logical example to leverage their advances : the human being. The tendency to make mistakes is thus an imperfection envied by machines. Current artificial intelligence systems are therefore capable of reproducing thoughts, a deductive mind and even reasoning in order to stock each piece of data into its own memory : an advanced learning process.

In the past machine translation software used to run with an algorithm splitting the whole text into fragments and then searching the meaning of these fragments in the memory. Depending on the language, the software would then adapt the structure of the fragments based on the grammatical rules of each language. But since the translation’s quality was not always perfect, online translation has recently been suited with a new advanced learning system, allowing the software to learn from its mistakes.

As an example, let’s take two languages intrinsically different : English and Japanese. Having encountered performance difficulties during a previous translation in this language combination, the software changes its strategy and chooses a “compromise” language. In this case, Korean acts as a bridge between two grammars that are too different. After a short analysis, the automated translation software first translates from Japanese to Korean, and then from Korean to English. It gets around the difficulty and improves the result.

By taking language as an example, we quickly understand how much modern AI learning, which imitates human reasoning, can become effective. We could think that the gap between AI and human beings is becoming smaller, but the non-mechanical mechanism of the human mind cannot be excluded from the equation. The nuances, the feelings, the cultural knowledge, etc. What makes humanity beautiful is the myriad of unsolvable enigmas that even the most powerful calculator,-let’s call it “computer”- could not solve.

Written by Gildas Mergny

Translated by Arthur Chevallier-Letort

InDesign : What if we Simplify our Lives ?

These are the words said by any translator who had to work on his document’s layout : bugs always occur at the last minute. And for good reasons, the files sent by customers are not always well thought out.

InDesign, a wonderful little desktop publishing software, allows you to automate the layout work very easily (according to those who use it). We will give you some very useful tips, your collaborators will love you and even adore you !


First of all, take the time to set up layers and templates. This may seem basic to some of you, but it is always good to remember it. These templates will provide a basic layout (automated page numbering, decorations, logos, etc.) that will not need to be systematically replaced.

Let’s then move on to paragraph styles. For translators, there is nothing better than clearly defined styles in the file. It should also be remembered that computer-assisted translation (CAT) software does not take into account style formatting, but only text. It is then necessary, during the conversion, to reapply the paragraph styles in the document. Styles also make it possible to take into account the difference in size of the translated segments. Indeed, from one language to another, the sentences will not have the same length even though they convey the same idea.

Right after paragraph styles, you have to deal with character styles. These are small nuances in a text, for example a word in italics, bold, underlined, or written in a different colour. To highlight them, we do not create a new paragraph style but a variant of the one used thanks to the character style.

Finally, a last very useful notion for having an optimized document is the implementation of a title numbering system. We could have presented it to you in paragraph styles, but the importance of the subject deserved its own part. A numbering of titles and parts allows you to better organize your document, and to have better references in case of bugs in the sequence of articles.

Many tutorials are available on the Internet, just ask the right question and forums, videos, websites and online PDFs will have the answer you are looking for. There is nothing complicated and translators will be pleased to work with you on your files.

So, what are you waiting for ?

Julie Daval

Translated by Arthur Chevallier-Letort

Sources : http ://

Online Resources for Translators

In the technological era that we live in, more and more online tools are available to translators. The advantage of these tools is that they do not require the installation of a computer program since the data is stored on the Cloud most of the time. Moreover, they enable different people to work on a project simultaneously, regardless of the time or place. But how do these online options differ from their offline alternatives ?


CAT Tools

There are several computer-assisted translation tools (CAT) that are fully available online. These tools, such as WordFast Anywhere or Memsource, are often free and more adequate than their free offline counterparts. Therefore, the features of an application like OmegaT can pale in comparison to the ones of their competitors. It is true that these tools possess a wide range of functionalities, but they often require in-depth knowledge and tedious manipulations.

On the other hand, major tools like MemoQ and SDL Trados are renowned for their power and, in terms of CAT, are celebrated worldwide. Some of their features, at first seen as accessories, have gradually become essential for the comfort they provide to translators. For instance, the autocomplete functionality might seem anecdotal, but it saves time and speed to translators who cannot work without it.

On their part, the CAT environments seem to be a decent compromise between accessibility and automated features. Being easier to use than Trados or MemoQ, but simpler than tools like OmegaT, online versions seem to be a good way for beginner translators to discover CAT tools. However, experienced translators might probably opt for alternatives that are more expensive yet more powerful.

Dictionaries and Termbases

The situation of dictionaries and termbases is far easier to identify. Regarding information, the Internet is a great and inexhaustible source. So yes, you can use it for your research, but you must remain watchful.

For translators, the internationalization caused by the Internet mostly affects the different varieties of languages. Thus, termbases such as TradooIT and Termium provide perfectly reliable information… for the people of Quebec. This issue can be found in many other languages : for example, differences between British English and American English, between Portuguese and Brazilian Portuguese or between Castilian Spanish and Rioplatense Spanish.

Thus, the main problem with online dictionaries and termbases is the source of information and its reliability. After all, choosing good sources is also one of the translators’ tasks.

David Loury

Translated by Sarah Deville

Source : http ://

The Revolution in Instant Translation

Everybody loves to travel and that is a fact. One of the major problems is the language barrier and it is widely known that when going abroad, you must first learn how to speak the language of the country or at least have a dictionary –  paper or digital – close at hand. However, times have changed and the use of these dictionaries is no longer necessary. Numerous groups of companies specialized in the field of technology and artificial intelligence have created new apps available on Smartphones and iPhones allowing an almost instantaneous speech and/or visual translation.


Could these new technologies possibly have a negative impact on the future of the profession of translators and interpreters ? Can we rely on technologies based on artificial intelligence and finally, what benefits do they offer ?

Two Different Jobs…

Translators and interpreters are different from one another, despite the consistent confusion between the two. Translators are given a certain amount of time to translate a source text into a target language. They usually work with computer tools to facilitate translation (CAT tools, translation memories, terminology databases) and can also carry out in-depth research on the subject or field of the source document in order to improve their translation result.

However, interpreters are asked to verbally translate a text almost instantaneously (a speech, for example) into the target language. There are three different types of interpreting :

  • Consecutive interpreting consists in writing down an interlocutor’s speech in order to deliver the same speech verbally to a third party’s language. In general, this type of interpreting requires the use of symbols as the interpreter does not have much time to write down every word and/or sentence.

  • Simultaneous interpreting consists in instantly delivering in the target language a speech that the interpreter is hearing in the source language. Generally, this service is commonly requested during conferences and meetings at the European Union or the United Nations. An interpreter, who can translate instantly only for twenty or thirty minutes, must work alongside another interpreter, who will take over at the end of the specified time. Each booth, therefore, houses two interpreters and is available in one foreign language.

  • Liaison interpreting is the process of instantly and verbally translating a speech into another language in order to help two people who speak two different languages understand each other. The interpreter is, therefore, the communication link between the two parties.

To summarize, a translator translates a written source text into their native language or into a target language, whereas the interpreter verbally adapts a speech in the source language into a target language which can be their native language or not. In both professions, the initial message of the source text must be communicated.

…A Similar Challenge

Despite differences between translators and interpreters, both may see their future threatened by the evolution of technology and artificial intelligence. Indeed, the market for linguistic services is increasingly exposed to technology, especially the recent development of new IT tools for translation assistance, which did not exist a few years back. However, nowadays, man and technology are working hand in hand and some studies believe that despite the constant evolution of technology, human translators cannot be replaced by machines.

Technology : New Apps and Software

Recently, new software and apps have emerged on the market, in particular on the various online application distribution platforms, such as App Store, Google Play. These apps offer a hybridization between translating and interpreting.

Have you ever tried to communicate with someone who does not speak your language ? For example, when a tourist asks you for information or directions ? How would you react ? Some companies specialized in technology and artificial intelligence have made these impossible communications, quite possible.

  • Almost instantaneous voice translation ? You just have to install the app on your smartphone. Once installed, you choose the language you want and speak into the microphone. The app will translate into the language of your choice and you can then let anyone listen to and/or read the translation. You can usually choose between approximately twenty foreign languages, but the rarest languages are not included in the app.

  • Almost instantaneous visual translation ? This process is the same as the previous one with a few exceptions. Using an app, instead of translating audio content, the app can translate visual content such as texts, menus, and even outdoor panels by using your camera. Thus, when you “videotape” a panel, the app can translate its content into the language that you have chosen.

With these new technologies, you no longer need to travel with large and heavy dictionaries. This is yet another improvement that breaks down the language barrier and facilitates communication.

However, despite a further expansion of technology in the language industry, artificial intelligence still cannot detect the nuances, contexts and human language idioms. This is why it is safe, for the time being, to say that technologies are not ready to replace linguists such as translators and interpreters. These programs, although somewhat limited and unreliable in some very specific areas (legal, financial, etc.), can be quite useful for some people, if not all people willing to learn languages.

Avatar Christelle Dilling

Source : https ://





Translated by Gildas Mergny