The old definition Wikipedia show for Machine Translation reads like this, “machine translation application is a program that attempts to translate text or speech from one natural language to another.” The new definition reads like this; Neural, AI, translation free code software is learning as fast as feeding with information to translate more accurately words, sentences or even documents.
Natural language definition shows the following “In neuropsychology, linguistics, and the philosophy of language, a natural language or ordinary language is any language that has evolved naturally in humans through use and repetition without conscious planning or premeditation.”
Words like application and attempts, are now substituted by software, evolve, super-use and super-repetition, unattached, development and execution and many more. Data is what fuels digital transformation, AI unlocks the value of that data, and hybrid multi-cloud is the platform to democratize the data. Today is not in off to request or not Machine Translation.
Most important software neural or not (seems that all of them are now depending on Neural and AI), are Google Translate, IBM Watson Language Translator, DeepL and Amazon Translate AWS, there are many others but far away to compete against those I mentioned. Google is a free platform poor for specific fields but fast and someway accurate for simple or standard content, Watson can offer a range of possibilities, arguing human translation is expensive but at the end, after all the “benefits” suggests that a professional human translator “must” review the end product. DeepL is even worse than Google and Amazon is almost the same as Watson (it cost 15 USD per 1 million words)…
A fast look of this shallow information make us go over the question about the use or not of Machine Translator. As a customer, it will be very convenient to have a robot translate my content if there is nothing complex or important that affects my product, as an ethical customer I will never use Machine Translation for my translation needs, as a translator the tools is becoming quite useful if I control the process to structure a readable content .
Invest in a Translation Machine software seems yet far for a need but the important access to help build the available software to translate more specific and specialized content. Translation machine market must become an open source to let linguist and other field professionals to take part and enhance products. As writing open code, translation machine should be a constant evolution of data developed collaboratively.
After the assessment of the different translation machines available using simple content or very specialized, results are quite different but no one is as precise as the human translation. I will recommend much more research, tools and sand atmospheres to obtain terminology, and linguistic structure to make sure it meets the diverse needs and expectations for everyone.
So far, translation machine is nothing but a tool to hasten together with other software and translation knowledge an excellent output. Many will differ and back up MT, but no one will show solid evidence ready to qualify TM as a reliable means of translating important content. Am I wrong?