In previous eras at year's end, I've offered Top 10 lists of things that might happen, three things to look forward to, or some other helpful list. This year I've decided to do something slightly different and a bit more personal.

I'd like to share with you the things that stood out to me during this past year among the many events and developments that happened in the world of translation. I've read other blog posts and comments that highlighted the large number of merger and acquisitions among LSPs. Clearly, that's relevant for a number of reasons, but it's not something that stands out to me and affects my work. (The only exception to that might be the purchase of Donnelley Language Solutions by SDL because it brings what may be the best corpus-based translation environment, MultiTrans, together with the most widely used translation-memory-based translation environment, SDL Trados. This offers interesting possibilities, but I think it's okay to get excited about it once we see what happens.) Or the seemingly sordid stories about TransPerfect's management. Or the recently unveiled considerations to possibly take Lionbridge public once again -- this time on the Australian stock market. That's interesting and scary at the same time (if one of the largest suppliers of translation-related services becomes a pawn in the hands of investors), but again not something that immediately and personally affects my life as a translator interested in the development of translation technology and the world of translation in general.

Without further ado, here are the developments that mattered to me this year (in no particular order):

• The rise of web-based translation environment tools
• Changes in privacy legislation and its impact on certain tools
• The rapid rise of the number of machine translation services
• Integration of voice recognition into translation environments
• How technology and services are linked (and our role in that)
• Diversity in the world of translation
• Community-driven approaches to form and find voices

Here we go:

It's not that so many new players entered this market last year, but the use of web-based translation environment tools has skyrocketed. In a recent survey by the Italian Way2Global in cooperation with translators' associations and universities (I will link to the survey once it's public), more than half of its respondents say they use web-based tools "often." I bet even a year ago that number would have been significantly lower, let alone three or five years ago.

It's important to note that there are not only positive aspects to web-based translation environment tools. There are also many different kinds of such tools. Let's start with the latter.

First are the web-based translation environment tools made to be only that, with no (or essentially no) desktop component. These include tools like Wordbee, XTM Cloud, Wordfast Anywhere, Memsource (which does have a desktop component that mimics the web-based interface), Smartcat, MateCat, Lilt, Transifex, Crowdin, Easyling, MotionPoint, Smartling, Lingotek, and even Google Translator Toolkit, as well as a number of tools owned and operated by language service providers. Some of the providers of these tools have seen the writing on the wall (or: in the cloud) for quite a while (Lingotek since 2006 and Google Translator Toolkit since 2009). While some of these providers offer options to use a private server for data storage and transfer, most of them use cloud-based servers with private partitions to separate the user's data from that of their teams (though in some cases the data can be used across users for analytical and some cross-training purposes).

A whole different kind of web-based tool are primarily desktop-based with an alternative web-based interface, including Trados, memoQ, and Across. In each case, the projected use case scenario is different. Trados, for example, seems to view their web-based tool, which is dramatically simplified in comparison to their desktop tool, more in terms of post-editing and less as a high-powered translation environment (in relation to the desktop tool). Across, on the other hand, has achieved complete feature parity between their desktop and web-based environments, while memoQ's web-based tool at least tries to play catch-up with the desktop-based older sister.

Why is this shift toward a web-based environment overall positive? For a number of reasons. First, these tools' design criteria typically differs from the desktop-based tools by being more streamlined and less overloaded with features. Clearly, people who see themselves as "power users" might not like that, but for them the desktop tools are still available. For the majority of translators, it's helpful to have a translation environment that allows them to be productive after just one or two hours of use.

Second, I think it's somewhat unrealistic to assume that the world of translation will be spared from a move to cloud-based environments. This simply matches an already existing reality.

And lastly, I'm delighted that this whole boring discussion about operating systems and devices has finally come to a close since web-based tools can clearly be used across the board.

The European Union's GDPR (General Data Protection Regulation) has created a lot of work for some and a lot of headaches for other translation professionals, but overall we've all benefited greatly. Case in point: Most publicly available generic machine translation providers (Google, Microsoft, DeepL) are now assuring users that their translation data will not be used for any purpose other than that user's translation if the paid API is used. That said, it's helpful to look into some other providers' fine print policies that are not quite so clear (Amazon, ModernMT, and others). In the case of neural machine translation, it is often -- correctly -- said that training data (unwittingly provided by you or your clients) cannot be reconstituted into the original texts, but it might still go against privacy agreements you have with clients to use their data as fodder for training MT engines.

The European data regulation has also helped other vendors of cloud-based tools (see above) to verify that their data usage adheres to the GDPR and is transparent.

While it still looks like the only way to make money with machine translation as a technology vendor is either to sell the technology as a whole (such as tauyou to TransPerfect, Safaba to Amazon, or AppTek to eBay) or to monetize it in convoluted ways a là Google and Microsoft, the MT market is getting more and more crowded. Crazily so! Just this past year, Amazon and IBM entered with new products; DeepL made lots of news (lately also with Portuguese, Russian, Japanese, and Chinese in the pipeline -- frustratingly available only with a European address and credit card if you want to use the paid Pro version); ModernMT is starting to make more noise (with an interesting -- but very expensive -- offer of an adaptive MT; more in the next Tool Box Journal); SDL Language Cloud has seemingly found its stride with a reasonable number of users; and Baidu, Sogou, Youdao, and Yandex have continued to come out with neural machine translation products that can be integrated into a translation environment (with or without the consideration of privacy concerns in these last few cases).

Then there are the many smaller companies that provide services to customize machine translation engines on the basis of their own technology (Omniscien, KantanMT, Systran, PangeaMT, etc.), and on top of that the many products that provide support for only one or two language combinations.

And aside from a direct, API-based connection to translation environments of the services mentioned above, tools like GT4T or RyCAT allow the bundling of some of these MT services and bring them into our translation environments.

In addition, there are now tools that aid the user in making decisions about which machine translation is to be preferred or about the presumed quality of a given machine-translated segment. The first of these companies was Fair Trade Translation, started in 2015 by two industry veterans (you can read about it here and here). The concept was interesting, but they were not able to make it work as a business. More high-powered than that is Intento (see here), which has made a bit of a splash this year. This tool allows the user to connect to a very large number of MT engines and evaluate how helpful the results are. And translation environment tool Memsource introduced an AI-driven service it calls MTQE (machine translation quality estimation -- see here) that uses Big Data from its network to judge whether machine translation matches can be deemed "perfect matches."

(As a side note, the use of artificial intelligence in translation technology aside from neural machine translation has been very interesting. At least three tools -- Memsource, Lilt, and Smartling -- have already introduced AI-driven features, and I know that in 12 months from now there'll be a lot more to add to that list.)

Back to MT: Many translators have found that machine translation can indeed provide benefits even for high-quality translation products in the form of an additional resource and often not as the main resource that needs to be post-edited. And I believe that's the most important thing in connection with machine translation that has happened in this past year. We have become more creative in how we use it and have realized that it requires an additional set of skills (to be able to deal with an additional batch of information), but overall many of us use it unapologetically for some of our project

How does this fit with a recent statement like this:

What true professional translator in his right mind, inside or outside of the EU, would want any machine translation in his translation environment? None indeed. DeepL's results may look great but is full of very risky flaws. No MT for me, no matter how good it claims to be.

Well, that might be true (or good for marketing) for some of us, but it's not for others. And this brings us to...

Hey memoQ!

Wouldn't it be amazing to translate in memoQ the same way you ask Siri for advice on a nearby restaurant? With the Hey memoQ app & memoQ 8.7, it is now possible! Turn your iPhone or iPad into memoQ's ear, and translate by dictating in more than 30 languages.

Learn more at http://bit.ly/heymemoq and download the app today.

In many recent talks, I've been using an image of a supernova as an illustration of what our world looks like. Not because we're at the end of an era like supernovae, but because it shows such a crazy assortment of pieces of a (former) whole.

We are diverse, and we can (and do) celebrate it. There is no reason to put down someone else's view of how translation is done properly as long as they are successfully making it as a professional translator. Yes, we do have a responsibility to present our profession to the outside world in a professional manner, but I think even that professionalism can be diverse. In that regard, the TranslationTalk Twitter account has been one of the highlights of my year and, I think, of the whole world of translation. While we did have some strongly opinionated curators, we treated each other with respect and recognized that we really are not competing against each other; instead, we're pulling together on one string that happens to be made out of many strands. I have been amazed at how little I've known about many of the strands that have been represented.

With some amusement I have been following Jochen Hummel's (the founder of Trados) campaign to announce the end of the CAT tool (aka translation environment tool) under the hashtag #sunsettingCAT. It's not quite clear to me whether this is Jochen's belated revenge on the new owners of "his" tool or whether he is just promoting his new company, but because of his prominence he's been getting a bit of publicity. You can read his ideas in this article. His argument is essentially as follows:

"The dramatic advances in Neural Machine Translation (NMT) have now made the whole product category [of CAT tools] obsolete. (...) Most players in the $50b translation industry, service providers but also customers, think that NMT is just another source for a translation proposal. In order to preserve their established way of delivery they pitch the concept of 'augmented translation.' However, if the machine translation is as good (or bad) as human translation, who would you have revise it, another translator or a subject matter expert?"

I prefer to categorize this line of argumentation from the person who most successfully marketed translation environments tools for many years as just another sign of how diverse we are. From the translators who view any use of machine translation as a sign of non-professionalism to the heralds of the end of translation (and quite literally "the end of translation" rather than the otherwise often-quoted "the end of translation as we knew it").

Technology has had a tremendous impact on how we conduct our business and practice translation. My preferred view -- and I know most of you would agree -- is that it's a business and a practice run by people, supported by technology. To what degree is up for anyone to decide.  

Imagine all the world's best dictionaries at your fingertips!

For a fixed monthly fee, on all your devices, integrated in your daily work applications!

WordFinder Unlimited -- One service, 5 applications and more than 300 dictionaries in 26 languages.

WordFinder Unlimited was voted the winner of the Process Innovation Challenge at LocWorld in Warsaw 2018.

Click here to find out 10 things you need to know about WordFinder Unlimited  

Over the last years a number of companies in the world of translation have focused on developing technology that merges aspects of workflow, professional translation, and machine translation into products that become the vehicle for delivering their company-specific services. Examples for this include MateCat, which offers free use of its translation environments and associated resources but pushes low-priced services along with every transaction, and companies like OneHour Translation or Unbabel, which offer highly transaction-based models that merge the above-mentioned aspects.

Smartcat approached the whole equation differently. They also offer free use of the translation environment, but they connect it with a comprehensive marketplace for the various players who offer and need services and charge a certain percentage of every transaction that occurs there.

I believe no other company or product has so publicly struggled and succeeded in fitting technology with users and services as Lilt. (One could argue that Lingotek has also gone through many different and comparable incarnations.)

In the course of just a few years, Lilt has gone from being a very translator-centric company that primarily focused on selling its products to translators, to a company that focused more on larger language service providers, to the company that now offers translation services alongside their technology and doesn't even give translators access to its technology unless they also, at least at some point, have sold their services through Lilt. I have mixed feelings about this. I greatly welcome Lilt's compensation for translators (based on an hourly fee) and the concept of high-quality services, but I also see this as a missed opportunity from the translator's viewpoint. In the past I've often written very positively about Lilt -- I like the technology and their new and creative approach to using it -- but I'm afraid that we (=translators) once again have been much too sluggish and passive in responding to a new and innovative product. Our passivity might have harmed ourselves again in the long run. Clearly it remains to be seen whether Lilt becomes the powerhouse its investors are hoping, in which case many translators will gain access to their technology (because they might render services to the service division). If not, we may have drawn the short stick.

This is one of my favorite things about technology, though: It makes for great stories, and it can become very personal.

Wordfast End-of-Year Promotions for Tool Box Journal Readers

�� Save $80 on your Wordfast Forward 2019 registration

�� Save €80 on the purchase of a new Wordfast Pro or Wordfast Classic license

These promotions end on December 31st. We wish you happy holidays and hope to see you next year in Martinique!

I don't have to repeat what I mentioned in the last Tool Box Journal about me having to rely more on voice recognition these days and the challenges and opportunities it brings. I was greatly excited this last year to see that voice recognition has become less exclusive. What used to be limited to the very few languages supported by Dragon (aka the company with the worst customer service ever) has now opened up to dozens and dozens of languages through voice apps by Google, Apple, and Nuance. The fact that memoQ now also offers the iOS app Hey memoQ as a memoQ-specific voice recognition tool for many languages is a really hopeful and positive sign that translation technology makers have gotten the message: Translators are diverse in the way they work, but united in their attempts to be productive.

And that brings us back to what might be a common denominator in much of this: community. I was greatly honored this year to be awarded an honorary membership by the ATA. I thanked the general assembly by saying that while our tribe is diverse and many of us are isolated in our offices, especially because of that I believe in community. For that reason it was particularly meaningful to be recognized by my community, and I promised to continue to pay it forward.

It's not easy to create and maintain community in a tribe like ours. How can it be? Our interests are wildly different, the languages we speak and work in are as diverse as the world itself, and we live very different lives. And yet we share a lot and have interests in common that we can articulate so much better when our voices are combined.

Translation Technology Wiki, which I hope will increasingly become a home where we can exchange ideas and, maybe more importantly, propose those ideas with a strong voice to the people who make the technology that has such an impact on us.

The other unifier that has been successful beyond my imagination is TranslationTalk. I mentioned how much I've benefited and learned from it, but that doesn't even begin to explain what this Twitter account has done for my pride and joy in my profession. Thank you all so much for being my tribe.

There's also the translation-specific URL shortener xl8.link. This has been a bit more of a struggle because of an attack by spammers that forced us to take the account off-line for a while and limit access. But we haven't lost any of the many shortened URLs and they should continue to be safe for the future. Rather than completely opening it up to all the nasty spammers again, we will be happy to supply you with a personal login so you can create unlimited shortened URLs and show the rest of the world that you're a translator and proud to be one.

Finally, on a very personal note, some of you know that I'm a Christian, and I'm thankful to be part of a fellowship of Christian translators that started this year and mirrors in many ways the diversity of the world of translation with Catholic, (soon) Orthodox, Mormon, and Protestant translators of many stripes.

All the best!

Service Release 1 for SDL Trados Studio 2019 has arrived

Service Release 1 (SR1) for SDL Trados Studio 2019 is packed full of exciting new features and improvements including:

• Enhanced alignment tool for better TM leverage
• Improved translation memory matches
• Simplified terminology management in the cloud
• Easier reviewing with enhanced Track Changes

Plus much more.

Visit our website to see what's new >>

The arrival of December provides perspective. We can pause and look back on another year quickly fading from present reality to past memory. As this year quickly draws to a close, here's a look at my top five develop

5. Productivity Software Designed for Interpreters -- This is not a new concept. Interpreters have been using software for years in different ways to prepare for jobs and increase productivity. But general-purpose software programs always leave interpreters wanting more, that one feature or function that would make life much easier when preparing or interpreting. And, let's face it, lead times are getting shorter, which means preparation time for assignments is also a scarce commodity. Right now, the development of interpreter-specific programs for tasks such as glossary management, term lookup in the booth (or courtroom, or examination room), terminology extraction, and research are beginning to mature. They have started to gain traction in the highly specialized but relatively small consumer base of tech-savvy, professional interpreters. Provided the economics work for these niche companies and interpreters make the financial and time investment needed to learn how to use these platforms, this development could be a boon for the wider profession. Standout platforms in this space include InterpretBank and Interpreter's Help. Two practitioner-researchers who are worth following in this space are Dr. Anja Ruetten (blog: blog.sprachmanagement.net) and Josh Goldsmith. Keep an eye on the development of automatic term extraction technology in the c

4. The Remote Interpreting Market Splits in Two -- 2018 was a banner year for remote interpreting. Remote interpreting software platforms continued to evolve, and the amount of remote work continued to increase, albeit nowhere at the speed that remote interpreting companies would like. (For more info on this, be sure to read my September 2018 column "What Kind of Interpreting Work Is There in the Cloud?" in the Tool Box Journal archives.) This year also showed a clear trend toward a bifurcation in the remote interpreting market between:

Remote interpreting that largely seeks to replace existing on-site interpreting delivery models to provide the service in venues that do not have room for interpreting booths (usually internal closed-circuit setups where interpreters are located nearby but in a different room) or to save money on travel costs.

Cloud-based interpreting that expands access to professional interpreters for meetings such as webinars, video conferences and conference calls, where meeting participants are also distributed geographically across a country or around the globe.

The former is "old market" on-site interpreting that is being disrupted by technology, the latter is "new market" made possible through technology. As professionals, we tend to focus more on old-market disruption. We would do well to pay more attention to and help develop new-market growth, which brings me to the next notable development...

3. Interpreters Begin to Make Peace with Remote Interpreting -- Since I started following this topic ten years ago, the generally-accepted position of the interpreting profession and just about every professional association regarding remote interpreting was that it shouldn't be done. Period. I think 2018 will be remembered as the tipping point for acceptance of remote simultaneous interpreting in its many different forms. The International Association of Conference Interpreters (AIIC) issued an updated position statement on "distance interpreting" recognizing that this kind of interpreting is taking place regularly and that an eventual official position regarding working conditions "must be based on independent and objective research" of which there is precious little. In the meantime, remote interpreting is becoming technically less complex and the need for simultaneous interpreting in new cloud-based meeting formats continues to grow.

2. The Power of 2.0 and 3.0 and 4.0 and... -- The lean startup philosophy that permeates Silicon Valley is having a growing effect on the interpreting world. This philosophy seeks to shorten product development cycles and allows technology companies to get a minimum viable product built (e.g. a cloud-based interpreting platform or a glossary management tool or a business model, for that matter) that may be rough around the edges but still gets the job done. They learn from initial user feedback (i.e. complaints and criticisms) and then incorporate improvements into the next iteration of the product and the next and the next. This philosophy also works for business models as well.

Sure, many interpreting technology startups have disappeared after being no more than a flash in the pan, but many others are no longer startups per se, but rather viable businesses that have been around for five years or more. If you compare their original product offerings or business models to what they offer today, the differences are striking. Here's an example, one remote interpreting company I have been tracking since its launch in 2014 has tried several different business models. It even rebranded its entire interpreting service offering, after deciding not to pursue traditional interpreting service contracts. Once they finally found, or rather built, their own niche, they have really taken off. The company's remote interpreting revenue for 2018 has already surpassed $1 million dollars.  

1. Artificial Intelligence Storms onto the Interpreting Stage and Stumbles -- The big development in the interpreting technology space was not a success but rather a series of embarrassing failures that highlight just how complex professional-level interpreting really is. In April 2018, Tencent's AI-powered translation engine was selected by the Boao Forum to provide real-time translation and subtitling from Chinese to English. The results were disastrous. (Read my April column on the debacle "It Finally Happened . . . AI Tried to Replace Conference Interpreters" in the TBJ archives.) Then in September, Chinese tech giant iFlytek was accused of using the work of professional conference interpreters and passing it off as the output of its own machine translation engine iFLYREC at the 2018 International Forum on Innovation and Emerging Industries Development in Shanghai.

These Artificial Intelligence (AI) moonshots and spectacular failures shed light on the attitude of many AI developers that there is nothing that the right algorithm can't solve. They also show a lack of understanding of the many factors that go into providing quality simultaneous interpretation, most importantly the human factors that are now being touted as the things AI can't do well or do at all-creativity, strategy and empathy, all of which are central to good interpreting.

All that said, AI will continue to encroach on the interpreting space. There will be more AI-powered speech-to-text and speech-to-speech translation devices, apps and services in the new year. But they shouldn't be dismissed. Successful interpreters will understand what sets them apart from AI and be good at articulating it to clients. To do that, we need to understand what this technology can and can't do. That way we can be perceived as the language experts we are and not nervous old-timers worried about the future.  

2019: There's No Going Back

A decade ago, smart devices of any kind were barely on the radar screen of practicing interpreters and service providers. Now, they are integrated, indispensable tools in many aspects of our professional lives. 2019 will see a continuation of that process. So, stay tuned, it's more important now than ever to stay abreast of technology developments impacting how we work.

Do you have a question about a specific technology? Or would you like to learn more about a specific interpreting platform, interpreter console or supporting technology? Send us an email at inquiry@interpretamerica.com.

If you would like to promote this journal by placing a link on your website, I will in turn mention your website in a future edition of the Tool Box Journal. Just paste the code you find here into the HTML code of your webpage, and the little icon that is displayed on that page with a link to my website will be displayed.

If you are subscribed to this journal with more than one email address, it would be great if you could unsubscribe redundant addresses through the links Constant Contact offers below.

Should you be interested in reprinting one of the articles in this journal for promotional purposes, please contact me for information about pricing.