Machine translation is not solved, but it's reached some surprisingly improved benchmarks for accuracy, so while it's a little presumptuous to call it solved, i

Machine translation is not solved, but it's reached some surprisingly improved benchmarks for accuracy, so while it's a little presumptuous to call it solved, it's not the most egregious exaggeration I've heard about machine learning this week.