Machines Just Beat Humans on a Stanford Reading Comprehension Test
AIs developed by Alibaba and Microsoft both outscored human participants.
Chinese retail giant Alibaba has developed an artificial intelligence model that’s managed to outdo human participants in a reading and comprehension test designed by Stanford University. The model scored 82.44, whereas humans recorded a score of 82.304.
The Stanford Question Answering Dataset is a set of 10,000 questions pertaining to some 500 Wikipedia articles. The answer to each question is a particular span of text from the corresponding piece of writing.
Alibaba claims that its accomplishment is the first time that humans have been outmatched on this particular test, according to a report from Bloomberg. Microsoft also managed a similar feat, scoring 82.650 — though, those results were finalized shortly after Alibaba’s.
Both Alibaba and Microsoft’s entries may have only beaten human participants by a whisker, but this is still a huge accomplishment when it comes to machine reading. As is already being explored, the technology will likely have many applications in areas such as automated customer service.
Machines are getting much better at accurately responding to queries by supplying the exact information that’s been requested from a large pool. As this action doesn’t necessarily require a human presence, we could soon see computers and AIs being tasked with this kind of responsibility.
In October 2017, Alibaba announced its plans to invest $15 billion in emerging technologies such as artificial intelligence, quantum computing, and the Internet of Things. The machine reading model is just one facet of a massive research initiative.
In less than 20 years, Alibaba has become one of the top ten largest companies in the world, primarily due to its success as an online retailer. The internet has changed the way that we shop, and as such the company is pumping money into research projects that will help ensure that it can keep up with the next game-changing advance in e-commerce.
Automated customer service that can remain active at all times, thus serving customers all around the world, is one such innovation. The technology could also be put to good use toward marketing, as AI could potentially be used to improve the accuracy of targeted advertising, or even as part of efforts to use social media to promote products.
China is becoming a hotspot for AI research, and Alibaba’s available resources will allow it to be at the forefront of this movement. Earlier this month, a $2.1 billion industrial park dedicated to the technology was announced for Beijing, as part of the country’s long-term goal of being the world leader in AI by 2030.
AI is already transforming the world we live in, and it’s only going to facilitate further change in years to come. The research being carried out today will have a profound impact on the world of tomorrow.