The true rise of the machines replacing people is far, far away as is the galaxy created by George Lucas. But that does not mean the translation/localization business is not endeavouring to replace a relatively expensive translator output with a much cheaper machine one. As the machines cannot think as humans yet, consider and pick the best word to be placed into each sentence, the results are often literally horrible, the output is often downgraded to a non-sensual computer code and the art of translation to a tasteless translated factory product.
I am certainly not one of those who generally refuse processing the machine translation (MT) output. I think that post-editing (as we call this service) is reasonable to be used for certain types of content only, it needs a large memory, legacy materials and terminology glossaries to “feed the MT engine”, and, of course, this engine must be a good one. It also heavily depends on a target language. I can assure you Czech (as any other Slavic language) with all the rules for declension, inflexion, conjugation, grammatical cases, etc. is quite a mouthful for the MT engine to chew and swallow without choking.
I generally have a good experience with IT content (e.g. software strings, user interface, documentation) where the terminology and clearly stated style help the MT engine providing a workable output. If this is the case, the productivity increases and the solution is beneficial and cost-effective for the customer. I would definitely recommend avoiding of MT post-editing efforts for marketing or similar areas that require creativity in order to catch the reader’s attention.
My outstanding analytical skills and general translation abilities allow me to find the part of the machine translated sentence that is misplaced, mistranslated or inappropriate in other respects. I will be more than happy to provide an assessment for a particular MT output sample and work on your projects when mutually beneficial for both sides.
Post-editing can be actually fun, but only when you can have a good laugh at the machine, wrongfully picking glossary terms that are totally out of context while the whole sentence makes more or less sense, not when the output reflects a non-sensual result and it’s necessary to start translating from scratch, that’s when the laughter vanishes in no time.