HKU Business School held its Inaugural ESG Conference

On 24 November 2023, the HKU Business School held its Inaugural ESG Conference, setting the stage for impactful discussions on cutting-edge research findings and recommendations for pressing environmental and social challenges.

It was a remarkable gathering of 75 brilliant academics, including faculty members and PhD students from a range of disciplines such as Accounting, Economics, Finance, IMM, Management, and Marketing, making it a truly interdisciplinary event.

Some of the thought-provoking topics covered included the use of AI as a teaching tool to bridge the gender inequality gap, the impact of institutional investors’ selection of green stocks on portfolio risk and diversification, and the effectiveness of incorporating environmental factors in management compensation to encourage “greener” behaviour.

Thanks to the organizers of this event, Prof. Roni Michaely, Prof. Shipeng Yan, and Prof. Guojun He, as well as all the incredible speakers and attendees, who voluntarily dedicated a whole day to academic discussion and the exchange of knowledge, their contributions have played an integral role in the success of the ESG research and made the event truly inspiring and insightful.

Presentation Topics:

Family Friendly Workplace Policies

Julian Costas-Fernandez, Sebastian Findeisen, Anna Raute and Uta Scönberg

Government Procurement and Corporate Commitment to Climate Change

Omri Even-Tov, Guoman She, Lynn Wang, and Detian Yang

Can Publicizing Information on the Carbon Intensity of Insurers’ Investments Affect Their Carbon-Intensive Investments?

Jeffrey Ng and Xiao Zhang

Deceptive Environmental Disclosures during Conference Calls

Shuqing Luo, Detian Yang, Zhige Yu and Guochang Zhang

Physical Climate Change Exposure and Pollution Outsourcing along Supplier Chain

Bingkun Zhang

Corporate Social Responsibility and Product Price: Evidence from the 2013 Indian Companies Act

Qingwei Wang, Jiang Zhang and Hong Zou

How green is green? Anatomy of ESG funds’ selection

Dunhong Jin, Roni Michaely and Menghan Wang

A BIT less of ESG scandals: How does bilateral investment treaties termination affect foreign multinationals’ ESG practices abroad?

Xin Chen and Yifei Zhang

Beyond the Bench: The Role of Climate Litigation in Averting Climate Crisis

Roni Michaely, Zoey Yiyuan Zhou, Joe Hong Zou

ESG-based CEO Compensation: Theory and Empirical Evidence

Jing Li, Chuan Lin, Ke Na and Guochang Zhang

Can Artificial Intelligence Improve Gender Equality?Evidence from a Natural Experiment

Zhengyang Bao, Difang Huang and Chen Lin

Inflation and Household ESG Preference

Wenzhi Ding, Chen Lin and Mingzhu Tai

When Minority Entrepreneurs Talk about Problems: A Study of Problematization Narratives in Pitches and the Entrepreneur’s Ethnicity

Soojin Oh

Biodiversity Risk and Bank Loans

Mengdie Deng and Tse-Chun Lin

Greenwashing of the Green Building Certification

Sumit Agarwal, Eduardo Araral, Mingxuan Fan, Yu Qin and Huanhuan Zheng

Other Events
港大經管學院最新多模態AI圖像生成能力排名出爐 部份中國人工智能模型表現突出
2025 | 學院新聞
港大經管學院最新多模態AI圖像生成能力排名出爐 部份中國人工智能模型表現突出
港大經管學院今日發表《人工智能模型圖像生成能力綜合評測報告》,針對15個「文生圖模型」及7個「多模態大語言模型」進行全面評估。研究顯示,字節跳動的即夢AI和豆包,以及百度的文心一言,在新圖像生成的內容質素及圖像修改的表現突出;而早前引起全球關注的DeepSeek最新推出的文生圖模型Janus-Pro,則在新圖像生成方面表現欠佳。研究亦發現部分文生圖模型雖然在內容質素方面表現優異,卻在安全與責任方面的表現強差人意。整體而言,與文生圖模型相比,多模態大語言模型整體表現較佳。
人工智能圖像生成評測成績單公布: 字節跳動百度表現亮眼,DeepSeek Janus-Pro表現欠佳
2025 | 研究
人工智能圖像生成評測成績單公布: 字節跳動百度表現亮眼,DeepSeek Janus-Pro表現欠佳
如今,人工智能領域的前沿模型技術已經從文本處理拓展至視覺信息的深度理解與生成。這些模型既能精准解讀圖像語義,又能根據文字描述創作出兼具真實感與藝術性的視覺內容,展現出令人驚嘆的跨模態理解與創作能力。本研究聚焦全新圖像的生成和基于現有圖像的圖像修改兩大核心任務,提出了一套系統性的人工智能模型圖像生成能力評測框架。我們基于多維測試集的構建與專家評審,對15個專業文生圖模型和7個多模態大語言模型的圖像生成能力進行了全面評估。結果顯示,字節跳動的即夢AI和豆包以及百度的文心一言在新圖像生成的內容質量與修改任務中表現突出,位列第一梯隊。對比不同類型的AI模型,我們發現,相對于專業文生圖模型,多模態大語言模型整體表現更佳。