Advertisement
Singapore markets close in 6 hours 9 minutes
  • Straits Times Index

    3,458.53
    -2.63 (-0.08%)
     
  • Nikkei

    39,508.84
    -85.55 (-0.22%)
     
  • Hang Seng

    17,406.56
    -62.80 (-0.36%)
     
  • FTSE 100

    8,167.37
    -31.41 (-0.38%)
     
  • Bitcoin USD

    65,692.71
    -1,922.34 (-2.84%)
     
  • CMC Crypto 200

    1,356.08
    -29.18 (-2.11%)
     
  • S&P 500

    5,555.74
    -8.67 (-0.16%)
     
  • Dow

    40,358.09
    -57.35 (-0.14%)
     
  • Nasdaq

    17,997.35
    -10.22 (-0.06%)
     
  • Gold

    2,413.60
    +6.30 (+0.26%)
     
  • Crude Oil

    77.26
    +0.30 (+0.39%)
     
  • 10-Yr Bond

    4.2390
    -0.0210 (-0.49%)
     
  • FTSE Bursa Malaysia

    1,620.79
    -8.89 (-0.55%)
     
  • Jakarta Composite Index

    7,304.46
    -9.40 (-0.13%)
     
  • PSE Index

    6,753.12
    0.00 (0.00%)
     

Chinese fashion student outsmarts AI teams to reach finals of Alibaba maths competition

A 17-year-old student majoring in fashion design, who is self-taught in maths, beat all the artificial intelligence (AI) teams to get into the finals of the Alibaba Global Mathematics Competition, according to the final list issued by the organiser, Alibaba's DAMO Academy.

Just over 800 competitors with an average age of 22 qualified for the final competition to be held on June 22. The majority of finalists are students, including those from Massachusetts Institute of Technology, Cambridge University as well as Peking University and Tsinghua University.

Jiang Ping, who studies fashion design at Jiangsu vocational school and spent two years studying maths on her own, became the surprise black horse. Jiang ranked 12th place in the final list and earned 93 points. She said she wants to prove "girls can also learn maths well".

Do you have questions about the biggest topics and trends from around the world? Get the answers with SCMP Knowledge, our new platform of curated content with explainers, FAQs, analyses and infographics brought to you by our award-winning team.

The winners of the final competition will share prizes of more than 4 million yuan (US$552,120).

This is the first time that Alibaba's open-to-all maths competition has allowed AI to assist with solving questions, and it attracted more than 563 teams from different universities and companies around the world. More than half of the participants were born after 2000, according to the DAMO Academy.

"AI challenges are a meaningful attempt, which can help us to break through the limitation of AI's capability," said Yin Wotao, the head of the decision intelligence lab of DAMO Academy and a committee member of the competition.

A signboard for DAMO Academy is pictured at an exhibition. Photo: Weibo alt=A signboard for DAMO Academy is pictured at an exhibition. Photo: Weibo>

All AI teams failed to get into the finals, with the highest score being just 34, DAMO Academy said. The top three AI teams were from Shanghai Jianping High School, Southwest Jiaotong University, and Renmin University of China.

Although the AI scores already reached the average level of human participants, they were still far from the level of a maths master, as the capability of the underlying model in maths was limited, according to the source from DAMO Academy.

AI can achieve basic scores through knowledge searching when facing solution and proof questions. However, it may trip up with an "illusion of knowledge" when the questions go further, especially for multiple choice, the source added.

However, the source pointed out that some participants used prompt words to make AI act as a "master of maths competition", and they even promised to give the AI US$300,000 if it could win the game, which increased the accuracy by 20 per cent.

The competition, held by Post owner Alibaba Group Holding, is open to all and aims to stimulate enthusiasm for mathematics.

This article originally appeared in the South China Morning Post (SCMP), the most authoritative voice reporting on China and Asia for more than a century. For more SCMP stories, please explore the SCMP app or visit the SCMP's Facebook and Twitter pages. Copyright © 2024 South China Morning Post Publishers Ltd. All rights reserved.

Copyright (c) 2024. South China Morning Post Publishers Ltd. All rights reserved.