Alibaba claims their artificial intelligence model has bested humans for the first time in the world’s premiere machine reading-comprehension test, paving the way for more advanced consumer facing AI.

The Chinese tech giant’s deep neural network model now owns the highest rank for the Stanford Question Answering Dataset (SQUAD) with a mark of 82.44 in Exact Match – providing exact answers to questions – beating the score by humans of 82.304.

They join Microsoft atop the leaderboard in joint first position. Microsoft has an exact match score above both Alibaba and humans, but Alibaba claim they were first to beat the human mark by one day.

SQUAD is a reading-comprehension dataset with over 100,000 question-answer pairs based on more than 500 Wikipedia articles. Every answer is found in a a text segment or span within the articles. Participants, which in the past has included Google, Facebook, IBM and Tencent, submit machine learning models which can analyse the dataset and provide exact match answers.

Consumer facing AI

According to Luo Si, Chief Scientist of Natural Language Processing at Alibaba iDST, the technology opens up several commercial and customer experience opportunities.

“We are especially excited because we believe the technology underneath can be gradually applied to numerous applications such as customer service, museum tutorials and online responses to medical inquiries from patients, decreasing the need for human input in an unprecedented way,” he said.

“We are thrilled to see NLP research has achieved significant progress over the year. We look forward to sharing our model-building methodology with the wider community and exporting the technology to our clients in the near future.”

The Alibaba technology was also used in their annual 11.11 mega sale to answer online customer queries during periods of high volume. During the 2017 event Alibaba processed US$25.3 billion of gross merchandise.

Australian organisations have used the Alibaba platform to successfully tap into the lucrative Chinese market.

Previous post

IDC launches digital transformation scorecard in APAC

Next post

Australia is now Amazon’s fastest growing marketplace: Analysis

Join the digital transformation discussion and sign up for the Which-50 Irregular Insights newsletter.