We often hear that artificial intelligence is important for economic growth, and although that statement has an intuitive meaning, there is not much hard data to back it up. However, a recent study by economists at MIT and the University of Washington in St. Louis offers some evidence that shows how AI tools increase commerce by allowing vendors to cross the language barrier.
Looking at the data extracted from eBay, the researchers compared sales between the Latin American countries of Spanish-speaking countries in the United States before and after the shopping platform introduced the translation driven by AI for the product lists in 2014. (Specifically, the translation tool affected listing titles and search queries, but not product descriptions.)
Before this, eBay offered automatic translation, but the use of AI significantly improved the accuracy of the service. Hopefully, a better translation will lead to higher sales, and that is exactly what the researchers found. His data showed that sales from the US UU The affected countries increased 10.9 percent after the tool was launched.
Researchers acknowledge that other variables may have caused this increase, but explained them in several ways. For example, they compared exports in this period to countries where languages are spoken that are not included in eBay's translation tools, and factors such as increased marketing spending and the length of title lists are controlled.
The researchers also compared this data with eBay's launch of AI. Translation tools for Russian, Italian and French. They found similar improvements and estimate that the introduction of AI translation between countries is equivalent to reducing their distance from each other (a factor that has a large impact on the volume of trade) by 26 percent.
"These comparisons suggest that trade The hindering effect of language barriers is of the highest order," the researchers write. "Improved machine translation has made the eBay world much more connected."
A blog post from the University of Washington in San Luis, the researchers said that their work could be used as a pessimism counterpoint on the stagnation of productivity rates around the world Erik Brynjolfsson, of MIT, a leading expert in productivity who is also co-author of this last article, has suggested that this could be due to the slow adoption of new technologies.
"The problem is that it usually takes time for organizations to increase complementary innovations, be they organizational or technological, to reap the benefits of AI, "said Xiang Hui, co-author of the article, to the University of Washington The Source This article by eBay, Hui says, shows that when technology is finally implemented can have significant effects.