LegCo uses AI to map debate priorities
AM730 · 4 SOURCESabout 7 hours ago5 MIN
Summary
Hong Kong’s Legislative Council Secretariat has for the first time used artificial intelligence to analyse all lawmakers’ speeches from a three-day combined debate on Hong Kong’s first five-year plan and this year’s Policy Address. Legislative Council President Starry Lee Wai-king (李慧琼) said all 88 lawmakers spoke, making 214 speeches over more than 26 hours, producing about 450,000 words of verbatim records. The AI-generated overview found that the debate segment on improving livelihood protection drew the most attention, accounting for nearly 30% of the overall speaking volume. The overview will be given to lawmakers and submitted to the government to help officials identify the debate’s main concerns more precisely.
Key Points
- The combined debate ran for three consecutive days last week and covered both Hong Kong’s first five-year plan and the motion of thanks on the annual Policy Address.
- Starry Lee said the Secretariat analysed speeches compiled with assistance from the smart transcription system “Zhishi Ting”, producing what she described as a more scientific overview
- Across six debate sessions, the livelihood protection segment had the highest participation and took up nearly three-tenths of total speaking content.
- The 10 most-mentioned keywords were talent, Northern Metropolis, innovation and technology, education, data/computing power, Greater Bay Area/Shenzhen-Hong Kong cooperation, AI, finance, youth and employment.
- According to the overview, 77 lawmakers mentioned talent and 71 discussed the Northern Metropolis, with suggestions on dynamic talent lists and differentiated university-town planning.
Why It Matters
The exercise gives the government a structured way to sort a very large volume of legislative feedback and may influence how policy priorities are refined after the debate. For Hong Kong readers, the findings also show that even in a debate framed around long-term planning, lawmakers devoted the largest share of attention to day-to-day livelihood issues.