Machine-based systems have beaten humans in the Stanford Question Answering Dataset (SquAD), for the first time, as Alibaba and Microsoft scored 0.1% better than the human benchmark. It’s a big step forward, and also indicative of the arms race between the Chinese web giants and their better-known western counterparts – but we’re still a very long way from a true AI, capable of general purpose intuition. The test is based on a collection of over 500 Wikipedia articles, which the test subject must study and then be quizzed on. The goal is to see whether the information in the articles has been understood correctly, with questions designed to test just how well that information has been ingested. Alibaba’s Deep Neural…