Last week, Alibaba put its newly developed AI model through comprehension and reading tests, and the results saw the machine bettering its human counterparts. The natural-language processing AI went through over 100,000 questions to combat the humans, and Alibaba claimed that it was the first time that a machine defeated people.
The test was staged at Stanford University, and the quiz is considered to be one of the most authoritative machine-reading gauges of the world. Alibaba’s deep neural network model was asked to give exact answers to the questions, and it finished with a score of 82.44. The rival humans trailed just behind at 82.304.
Although Microsoft’s AI scored 82.650 on the same quiz, Alibaba reported that this result was concluded after their machine’s, which makes the latter achieve the pioneering honor.
Alibaba AI model was developed by the company’s own midst. The Chinese titan followed Tencent and Baidu in developing AI to improve social media feeds, target ads and services, and autonomous driving. This is an evident part of Beijing’s endorsement of the technology to take it to a national-level plan: to lead the entire nation to be the leader in the industry by 2030.
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. – Luo Si, the chief scientist for natural language processing at the Alibaba-owned institute
The natural-language processing system is able to mimic the comprehending of linguistics of the humans. Stanford designs the set of questions based on over 500 Wikipedia articles and even ensures that the machine-learning models are tested in terms of their abilities in processing big scale data before providing precise answers.