I seem to come away from every translation conference with the idea of one dominant theme. At the ITI conference last weekend the dominant theme seemed to be Machine Translation.
Not to be confused with Translation Memory (TM), which is already extensively used in the profession, Machine Translation is computer generated translation based on statistical analysis.
In theory Machine Translation will speed up the translation process and reduce costs. It would allow a rough and ready translation to be quickly generated for post-editing by human translators. The companies promoting the software for doing this showed examples of light touch post editing. In this case blatant errors and mistranslations are edited but the finer points are ignored provided the text can be understood. For my part I think it would be very difficult to apply a light touch.
While the impression was given that Machine Translation was increasingly going to figure in the translation market, it was also certain that it would not compete with human translators, as it would appeal to a different sector of the market.