Just Plain Mean
- May 18, 1997
I know all you guys out there at UC Berkeley are nodding your heads saying, "How could AI NOT be smarter than a Stanford grad?" but I digress. The basis of the headline statement lies in a Stanford University Reading and Comprehension testing database. AI technologies from both Alibaba and Microsoft squared off against a 100,000 question dataset. Alibaba scored 82.44 and Microsoft scored a 82.65 a day later, no doubt resting on the AI research done in the previous years with Clippy. "Rival humans" achieved a 82.304. This all sounds well till you realize the datasets are based on 500 Wikipedia articles, so now at least we know that the AI technologies actually got no smarter at all.
So-called natural language processing mimics human comprehension of words and sentences. Based on more than 500 Wikipedia articles, Stanford’s set of questions are designed to tease out whether machine-learning models can process large amounts of information before supplying precise answers to queries.