Think back to those pesky reading comprehension tests you used to take in school. Trust me, they are still on the SAT and still challenging. At least, for us carbon-based intelligences. From the rapidly developing artificial intelligence perspective, we just got poned.
Not one, but already two, artificial intelligence natural language processing algorithms outscored the average person's score on a rigorous reading comprehension test designed by Stanford researchers. The first to claim the honor of "better than people" was Alibaba's artificial intelligence. Less than a day later, Microsoft's artificial intelligence scored above Alibaba's. Both scored above humanity--and neither are in the lab.
What does artificial intelligence read well?
Anything. The Stanford researchers did a sizzling job of making sure there was variety of mostly-factual material for the test-taker to show off their understanding of situation. Their paper reads:
"To retrieve high-quality articles, we used Project Nayuki's Wikipedia's internal PageRanks to obtain the top 10000 articles of English Wikipedia, from which we sampled 536 articles uniformly at random. From each of these articles, we extracted individual paragraphs, stripping away images, figures, tables, and discarding paragraphs shorter than 500 characters. The result was 23,215 paragraphs for the 536 articles covering a wide range of topics, from musical celebrities to abstract concepts."
Jamie Condlife, senior tech editor at MIT Review, adds some context in The Download. "This isn't comprehension the way humans think of it. It's neat, but the AI doesn't really understand what it reads--it doesn't know what 'British rock group Coldplay' really is, besides it being the answer to the Superbowl question. And there are far harder language problems that humans still beat computers at," he shared.
A real business case for artificial intelligence
These artificial intelligences may not be "working hard" on the fine points of human perspective, but they sure are hard at work. Specifically, Alibaba has its language processing AI at work in sales, answering 95% of questions from customers about products at peak times. This approach helped ring in $25 billion in a single day--Single's Day--a peak Chinese shopping day, according to a company press release. It was a 39% increase from the previous year when people did the work.
"The first technology revolution caused World War 1. The second technology revolution caused World War 2. This is the third technology revolution," Alibaba CEO Jack Ma shared with CNBC, a few months before his artificial intelligence became the first to outperform people in English language reading. "People are already unhappy because machine learning kills a lot of jobs. The way to figure out the job creation, one of the best ways, is to figure out how to help small business sell their local products across the board. We have to prepare now, because the next 30 years will be painful."
Let's be clear--Ma has no problem with machines being smarter than us. He pointed out that we've made busses and trains faster than us--why not make machines smarter as well as stronger? He does want to ease the pain of the labor and lifestyle transition he sees ahead, which also includes some great things, like working less. "Wisdom is from the heart; machine intelligence is about the brain. You can always make a machine to learn the knowledge, but it's difficult for a machine to have a human heart," he put it.
"We have to repair the roof, while the sun is still shining," he suggested, referring to the fast changes needed from government and business to prepare society for a changing world.