ROME--(BUSINESS WIRE)--For the first time in history, Translated, the leading global language services provider and pioneer of AI-powered localization services, quantified the speed at which we are approaching the singularity in artificial intelligence (AI). The discovery was made possible by analyzing data from 2 billion post-edits collected from several years of client translations. CEO Marco Trombetti previewed the research during the Association for Machine Translation in the Americas 2022 conference, where he was invited to present the opening keynote speech.
Many AI researchers believe that solving the language translation challenge is equivalent to producing Artificial General Intelligence (AGI). The evidence Translated has provided about reducing the gap between what expert human translators and an optimized machine translation system can produce is quite possibly the most compelling evidence of success at scale seen in both the MT and AI community.
If the progress in machine translation (MT) quality continues at its current rate, in about six years the highest-performing professional translators will spend the same amount of time correcting a translation produced by MT as they will correcting one completed by their peers. That is what we consider the singularity in translation. The exact date when that will happen could vary somewhat, but the trend is evident.
Translated’s finding relies on data representing a concrete sample of the translation production demand. It consists of records of the time taken to edit over 2 billion MT suggestions by tens of thousands of professional translators worldwide working across multiple subject domains, ranging from literature to technical translation. It also includes fields in which MT is still struggling, such as speech transcription.
Over the years, Translated has continually worked to evaluate and monitor machine translation quality. In 2011, the company standardized its methodology and settled on a metric we call “Time to Edit,” the time required by the world’s highest-performing professional translators to check and correct MT-suggested translations. Time to edit is considered the best measure of translation quality because there is no concrete way to define this metric other than measuring the average time required to check and correct a translation in a real working scenario. By switching from automated estimates to measurements of human cognitive effort, Time to Edit reassigns the quality evaluation to those traditionally in charge of the task: professional translators.
Full report on Translated’s website
No comments:
Post a Comment