Back to “human”

by Vasso Pouli, CEO at Commit

This is not a usual newsletter entry, it will not be about a specific localization subject nor a very informative or educational one. It is more of a ‘putting-my-thoughts-to-writing’ piece which however are open for questioning and debate by you, the readers.

There were two instances that sparked this idea, a funny one and a rather sad one; but let me start with the funny one and see whether we will get to the sad one.

For those of you who don’t know me, I have a rather strong-minded but cute 3-year-old daughter whose Christmas gift request was ‘Chatty’, a robot penguin. Well, Chatty has a pre-installed set of instructions which it acknowledges using voice recognition SW and acts accordingly. Chatty is a nice penguin with very funny responses and pre-recorded messages and every time we turned it on, we had a blast, but that was only as long as an adult voiced the instructions, i.e. when Chatty said ‘I am hungry’, we said ‘Eat’ and then there were a bunch of funny chewing sounds and Chatty thanking us for its delicious dinner and asking for a drink, and you can imagine how the story goes… But everything fell apart when my daughter wanted to boss it around being the one to be voicing the orders. At the age of 3, well… let’s just say that she has not yet mastered the art of clear articulation, and instead of ‘Stop’ she uttered ‘thtop’ or thought that it would be nice of her to add its name in the question i.e. ‘How are you, Chatty?’ or be even nicer and ask politely by adding ‘please’ in the order. Chatty was having trouble understanding her and either said, very successfully I may add, that she did not make sense or that she had not eaten her food and she did not speak loud or clear enough. At first, she was troubled and insisted, but then she was concerned, then angry because she could not understand why being polite did not have the expected outcome in this case nor what she was saying wrong when she uttered ‘thtop’ instead of ‘stop’. So now poor Chatty sits deactivated on the top shelf of her drawer wishing, I imagine, it could interact more or better with her, as this would be more fun. Right, Chatty…?

And you are probably wondering why I am saying all this… Well, we are in the era of Amazon deliveries with drones and instant service, we interact with machines all the more often, probably much more than we do with people, and we have a reason for doing so; it is easier, it is transactional, it is quick, so quick we have no time to actually mentally process it, and this makes it automatic. And automation is good, it saves us time, it frees our hands from the mundane and procedural tasks supposedly allowing us to deal with the more creative and challenging ones. But are we up to the task? This would require inspiration, interpersonal interaction and cooperation. How can we achieve this when we expect humans to communicate in a pre-conditioned manner and we condition both our words and our actions in order to trigger the expected outcome?

At the end of the day, when neither of the recorded options in a voice service suits our question, when our instantly generated invoice has the wrong information, when we accidentally press the wrong button and everything crashes, when we want a cafe latte with double shot of espresso, low fat milk, Stevia instead of sugar and a straw -even though it is  a hot drink-, we need to talk to a person, to an actual human being. Only they can be insightful, flexible, creative and add meaning and value to our interaction, and only then can we be genuine, can we be ourselves, even if we may not utter clearly enough or even if we are more verbose or polite than what would be expected.

We are all striving for a leaner, more automated and streamlined workflow, but what if we strived for more meaningful human interactions, where every ‘ping’ has its equal and corresponding ‘pong’ instead of the same mundane stereotypical preconditioned automated ‘ping’, ‘ping’, ‘ping’.

And why don’t we leave the sad story for another time?

What do you think?

Does technology threaten translation?

by Dina Kessaniotou, Project Coordinator at Commit

The question of whether technology threatens translation depends on many different factors, and, basically, on how people conceive its purpose. The answer is indissolubly tied to how effectively the involved parties can leverage its advantages, identify its disadvantages and set the limits.

If we take a step back and consider when technology first impacted the translation process, we will be able to see the big benefits it has brought about: the broad use of the Internet, even if not a translation-specific tool, resulted to a tremendous change in translation, compared to the old-fashioned, paper-based ways, in terms of quantity, speed and quality in search. There was an exponential increase in the volumes of information available to linguists. Search became much easier and much more effective, as huge amounts of data, with multiple possibilities of customization, were made instantly and directly accessible.

At a later stage, translation-specific technology, namely the CAT tools, offered many valuable advantages to all players of the translation production cycle. Linguists were able to accumulate their knowledge and previous research effort, store it and organize it in a way they could easily retrieve and reuse it in the future. They could therefore eliminate repetitive work, increase their speed, reduce turnaround times for their clients and, of course, keep consistency – one of the most painful tasks in many types of content. This means savings in time and costs for all involved parties in the translation process. This also means ability to focus on brand new content. As there are still huge volumes of untranslated content, clients will normally be more willing to push this content into production in the near future. This is what happened during the last few years and will probably continue to happen, given that major providers in different fields of specialization have realized the importance of localization. Based on the above, translation technology is clearly a faster way to growth.

The natural evolution of CAT tools was the development of Machine Translation systems. At this point, and especially at the earliest stages of the development, the usefulness of technology started to be questioned. Many linguists thought it really threatened translation as it aimed to replace the human brain. In fact, there is still no machine that can catch all nuances and intangible elements of a language and adapt them in a different language. Even the more “flat” texts evoke specific feelings and emotions that should be properly conceived and transferred. So there is no way for machines to “threaten” translation. This doesn’t mean that the MT technology can’t be fruitful, especially if linguists are constantly involved in the MT development. Instead of being skeptical about machines, we should rather make them work for us. What might need to change is the way linguists offer their knowledge. Some years ago, it might have been difficult to perceive how CAT tools would increase efficiency and profitability, not only for clients but also for linguists. Nowadays, a considerable part of linguists cannot imagine their lives without tools.

The fundamental purpose of technology is to be continuously aligned with the challenges of the market and contribute dynamically to the linguists’ efforts for high quality services – and not just to cut costs by delivering automated results in one step. The only thing that can downgrade its usefulness is the lack of understanding of its real mission. When used effectively, technology can bring exclusively positive results and is really a valuable and profitable investment for all involved parties.