AI衝擊環境 碳排放有何轉機

今時今日，人工智能（AI）技術突飛猛進，從智能助手到自動駕駛，從工業生產到醫療診斷，AI的應用已深入人類生活的方方面面。根據國際數據公司（International Data Corporation）預測，全球AI市場規模將從2022年的1324億美元增長到2027年的5124億美元。

在為AI帶來的創新便利歡呼之餘，社會各界是否意識到，這場科技革命正在悄然對地球環境造成巨大影響？事實上，AI發展過程中產生的碳排放問題，已經到了不容忽視的地步。

隱形殺手：AI訓練的碳足跡

要理解AI對環境的影響，必須揭開AI模型訓練的面紗。現代AI模型，尤其是大型語言模型的訓練過程，需要海量的數據支持和龐大的計算資源。根據馬薩諸塞大學阿默斯特分校的最新研究，訓練一個大型AI模型產生的碳排放量高達62.6萬磅，相當於5輛汽車從生產到報廢的全生命周期碳排放總和。

具體而言，GPT-3的訓練過程，大約產生552噸二氧化碳；至於規模更大的GPT-4，碳排放量估計超過1000噸。令人尤其擔憂的是，這個數字正在不斷攀升。在「大模型即王道」的行業共識下，各大科技公司爭相發展更大規模的模型，導致能源消耗呈指數級增長。到2030年，估計AI行業的碳排放量將佔全球碳排放的3.5%。

數據中心：AI時代的耗能巨獸

當前AI大模型的能源消耗令人咋舌。史丹福AI研究所數據顯示，GPT-3單次訓練耗電1287兆瓦時，相當於3000輛特斯拉電動車各行駛20萬英里的總耗電量，產生552噸二氧化碳排放。

從日常應用看，ChatGPT回應用戶1次，耗電量達2.96瓦時，差不多是標準谷歌搜索（0.3瓦時）的10倍；由AI驅動的谷歌搜索更需8.9瓦時。水資源消耗同樣驚人：GPT-3訓練期間，耗水近700噸，每20至50個問題就需500毫升水。2022年，Meta公司僅在數據中心冷卻上，就已耗水260多萬立方米。

能耗激增的根源

人工智能大模型的高能源消耗主要源於兩大核心因素。首先，隨着AI技術的快速反覆運算，晶片需求量急升，直接推高了電力消耗。現代AI模型的訓練和推理過程使用大量計算資源，這些資源主要依賴高性能硬體，包括圖形處理單元（GPU）和專用集成電路（ASIC），運行時耗電極大，以應付複雜的計算。隨着AI模型規模不斷擴大，所需的計算能力也呈指數級增長，對高性能晶片的需求亦有增無已，令整體電力消耗量飆升。

再者，AI模型訓練過程需要強大的算力支持，這些運轉不息的數據中心產生高熱量，必須冷卻處理。中心作為AI計算的核心基礎設施，其能耗問題尤為明顯。伺服器和儲存設備在高負荷運行時會釋出巨大熱量，如果不及時散熱，設備的性能和壽命都會嚴重受損。因此，數據中心都配備高效的冷卻系統，以確保設備在適當溫度下操作。

數據中心營運成本結構中，電費佔總成本的60%，其中超過40%用於冷卻散熱系統；特別是在風冷數據中心，超過60%的電力用來散熱，真正用於計算的電力還不到40%。這種能源利用效率的失衡，導致全球數據中心的耗電量已達10年前的近10倍。傳統的風冷系統雖然成本較低，但效能亦較低，難以滿足現代數據中心的高效散熱所需。相比之下，液冷系統雖然初期投資較高，但散熱效率亦高，能夠大幅降低數據中心的能源消耗。

此外，中心的選址和設計也對能耗有重大影響。不少數據中心建在電力成本較低而氣候炎熱的地區，加重冷卻系統的負擔。為了提高能源利用效率，數據中心應優先選擇氣候涼爽、能源供應穩定的地區，並採用模組化設計，以便根據需求靈活調整計算資源的配置。

最後，AI模型的訓練和推理過程還涉及大量的數據傳輸和儲存，這也增加了能源消耗。隨着數據量持續增加，中心因需要更多儲存設備和更高頻寬才足以應付，而進一步推高能耗。由此可見，企業應積極採用數據壓縮和優化傳輸技術，減少不必要的程序，以便降低能耗。

企業解決方案和政策建議

面對AI技術帶來的環境挑戰，企業和政策制定者需要採取一系列減碳措施。一、企業應積極採用綠色能源和節能技術。通過投資太陽能、風能等可再生能源，減少對傳統化石燃料的依賴。二、企業應優化AI模型的訓練演算法，減少不必要的計算量，從源頭上降低能源消耗。三、企業應加強數據中心的管理和技術升級。採用液冷等高效冷卻技術，提升能源使用效率。通過智能調度和負載均衡，減少數據中心的閒置資源浪費。四、企業可通過虛擬化技術，整合計算資源，進一步降低能耗。

在政策層面，一、政府應制定嚴格的能效標準，推動AI技術的綠色發展。通過稅務優惠和資助，鼓勵企業採用節能技術和可再生能源。二、加強對數據中心的監管，制定能源效益評估標準，推動行業整體能效提升。三、政府和業界攜手推動公眾和企業的環保意識。通過碳排放交易、碳抵消等機制，減少AI技術對環境的負面影響。教育和宣傳是不可或缺的範疇，只有當AI發展與環境保護之間的關係廣為人知，才能形成社會共識，合力應對。

危機中的曙光

AI技術勢不可擋，但我們必須確保其發展不以犧牲環境為代價。通過科技創新、企業自律、政府引導和社會監督等多管齊下的方式，在享受AI便利的同時，亦可把環境影響降至最低。根據國際可再生能源機構的預測，若能採取積極措施，到2030年，AI行業的碳排放增長率可以控制在每年5%以內。

作為這個時代的見證者和參與者，我們每個人都應該關注AI發展帶來的環境問題，用實際行動支持綠色AI的發展。只有這樣，才能夠確保AI技術真正造福人類，而不會變成地球的另一個負擔。在追求科技進步的道路上，環境保護絕不應作為可有可無的點綴，反之應是必須嚴格遵守的底線。讓社會上下齊心協力，推動AI走向更加綠色、可持續的未來。

通過政策引導、科技創新和公眾參與，香港不難在2035年之前，實現AI行業的碳中和目標，為全球可持續發展作出貢獻。

本欄逢周三刊登

章逸飛博士
港大經管學院經濟學高級講師

Dr Yifei Zhang

11 December 2024

Nowadays, with the advancement of artificial intelligence (AI) technology in leaps and bounds, AI applications have permeated various aspects of human life—not only from smart assistants to autonomous driving technologies but also from industrial production to medical diagnosis. According to the International Data Corporation, the global AI market value is expected to rise from US$132.4 billion in 2022 to US$512.4 in 2027.

While the convenience of AI innovation is applauded by all sectors of society, does it also raise the community’s awareness that the technological revolution is subtly exerting a tremendous impact on the global environment? As a matter of fact, the problem of carbon emissions arising from the AI development process has reached such a state that it can no longer be ignored.

The invisible killer: the carbon footprint of AI training

To understand the impact of AI on the environment, it is necessary to unveil the true face of AI training models. The training process for modern AI models, particularly large language models, requires massive amounts of data and calculation resources. The latest research by the University of Massachusetts Amherst indicates that carbon emissions from training a large AI model can reach 626,000 pounds, equivalent to the total emissions from five vehicles throughout their entire life cycle, from production to disposal.

Specifically, approximately 552 tonnes of carbon dioxide are emitted during the training process of GPT-3 while the CO2 emitted from training the even larger model, GPT-4, is estimated to exceed 1,000 tonnes. Of particular concern is that these figures continue to go up. Under the sectoral consensus that “large models are the order of the day”, giant technology companies have been vying to develop even larger models, resulting in exponential surge in energy consumption. The AI sector’s carbon emissions are forecast to account for 3.5% of the world’s total carbon emissions by 2030.

Data centres: an energy-guzzling beast in the AI era

The energy consumption of large AI models has now reached an alarming level. Data of the Stanford AI Laboratory shows that one single training session of GPT-3 typically uses 1,287 megawatt-hours of electricity, equivalent to all the power consumption of 3,000 Tesla electric cars each travelling 200,000 miles, emitting a total of 552 tonnes of carbon dioxide.

In daily use, every response generated by ChatGPT requires 2.96 watt-hours of electricity, almost 10 times that (0.3 watt-hour) for a standard Google search. Each Google search powered by AI even utilizes 8.9 watt-hours. The water resource consumption level is also alarming. During its training, GPT-3 consumes close to 700 tonnes of water. For every 20 to 50 questions, 500 millilitres of water are required. For cooling of its data centres alone, Meta used over 2.6 million cubic metres of water in 2022.

Root causes of escalating energy consumption

The colossal energy consumption of large AI models can mainly be attributed to two core factors. First, the rapid iterations of AI technology have significantly stimulated the demand for chips, directly pushing up electricity consumption. The training and inference processes of modern AI models deploy enormous computational resources, which primarily rely on high-performance hardware, including graphics processing units and application-specific integrated circuits. This hardware is highly energy-intensive when running complex computations. As AI models keep expanding in size, their computational capabilities have seen exponential growth, resulting in an ever-increasing demand for high-performance chips and, consequently, mounting energy consumption.

Furthermore, substantial computational power is essential for supporting the AI model training process. The around-the-clock data centres generate excessive heat, necessitating cooling treatments. Energy consumption is an especially severe issue for data centres, which serve as the core infrastructure for AI computation. Servers and storage devices running at high loads release a vast amount of heat. If the heat is not dissipated in time, both the performance and lifespan of the devices will be seriously compromised. Hence, data centres are equipped with super-efficient cooling systems to ensure that the devices operate at optimal temperatures.

In the operating cost structure of a data centre, electricity tariffs account for 60% of the total cost, of which over 40% is spent on cooling systems. At an air-cooling data centre in particular, more than 60% of electricity is used for cooling while less than 40% is used for computation. As a result of this energy utilization imbalance, the energy consumption of data centres around the world is now almost 10 times more than it was a decade ago. Traditional air-cooling systems are less costly but also less efficient, making them incapable of meeting the requirements for high-efficiency cooling. In comparison, except for a large-scale investment at the initial stage, liquid-cooling systems are more efficient, thus sharply reducing energy consumption at data centres.

In addition, the site selection and design of data centres have a significant impact on energy consumption. Many data centres are located in areas with lower electricity costs but in hot climates, placing a heavier burden on the cooling systems. To enhance energy utilization efficiency, priority should be given to locations with cooler temperatures and a stable energy supply. Besides, a modular design should be adopted so that resource allocation can be flexibly adjusted according to needs.

Finally, the training and inference processes of AI models also involve huge amounts of data transmission and storage, which inevitably boost energy consumption. As more and more data is created, data centres need extra storage devices and greater bandwidth to cope, which further expands energy consumption. These facts demonstrate that companies should make use of data compression and transmission optimization technologies to cut down energy consumption by minimizing unnecessary procedures.

Corporate solutions and policy suggestions

In the face of the environmental challenges from AI technology, companies and policy-makers need to take a series of carbon-reduction measures. First, businesses should maximize the use of green energy and energy-saving technologies. Investments should be made in renewable energy sources such as solar energy and wind energy to minimize reliance on traditional fossil fuels. Second, enterprises should optimize AI model training algorithm to streamline computation, cutting down energy consumption at source. Third, they should enhance data centre management and upgrade technologies; use high-efficiency solutions such as liquid-cooling to promote energy utilization efficiency; and minimize waste of idle resources through smart dispatch and load balancing. Fourth, through virtualization technology, companies can integrate computational resources to lower energy consumption.

In terms of policy-making, the government should first set strict energy efficiency standards and promote green development of AI technology; and, through tax concessions and funding, encourage enterprises to adopt energy-saving technologies and renewable energy sources. Second, regulation of data centres should be strengthened and energy efficiency evaluation standards should be established to promote overall energy efficiency. Third, governments and industry should join hands to spread environmental awareness among the public and businesses. The negative impact of AI technology on the environment can be minimized through such mechanisms as carbon trading and carbon offsets. Both education and publicity are indispensable, as only when the relationship between AI advancement and environmental protection is widely known can a social consensus be reached and concerted efforts be made to address the problems.

Glimmers of hope amid crisis

The trend of AI technology may well be overwhelming, but we must ensure that the environment will not be harmed as a result. Through technological innovation, corporate self-regulation, government guidance, and social oversight, environmental impact can be minimized while the convenience of AI can be enjoyed by all. The International Renewable Energy Agency predicts that, with proactive measures, the annual growth in carbon emissions by the AI industry can be controlled within 5% by 2030.

As witnesses and participants of this era, each and every one of us should be concerned about the environmental issues brought about by advancements in AI and take concrete actions to support its green development. Only through this approach can we ensure that the AI technology benefits mankind instead of becoming another burden on the Earth. In our quest for technological breakthroughs, environmental protection should be the bottom line that must be upheld, not just a token gesture. Let all sectors of the community make concerted efforts to drive AI towards a greener and more sustainable future.

Through policy guidance, technological innovation, and public engagement, the path will be paved for Hong Kong to achieve the AI industry’s carbon-neutral goals by 2035 and contribute to the sustainable development of the world.

當AI敲門 音樂產業命運何去何從?

今天當我們戴上耳機聆聽音樂時，會否意識到超過30%的音樂已經是由AI生成的？從去年一首AI生成歌曲Heart on My Sleeve在Spotify上一舉獲得逾2000萬的點擊，到今年年初，最大唱片公司環球唱片（Universal Music）因為AI音樂泛濫，而把旗下所有音樂從最大手機短視頻平台TikTok上撤銷，再到今年年中環球唱片和TikTok達成協議，後者同意在所有AI音樂視頻上加上「AI音樂」的標籤，AI技術對音樂產業的影響來勢洶洶。

是喜還是憂？

隨着AI技術的發展，各大科技公司的觸角已經深入到音樂行業。比如，Meta公司去年推出的MusicGen模型可以根據用戶輸入的文字生成音樂，而另一家公司Stability AI在今年推出的Stable Audio 2.0模型更允許用戶上傳現成的音樂，從而生成風格迥異的新音樂，音效甚至堪比黑膠唱片。

喜的是，AI讓非音樂專業出身的普通人不僅可以輕輕鬆鬆「創作」音樂，甚至還可以利用這些「創作」獲得額外的收入。美國一家初創公司Boomy允許用戶將其利用AI模型生成的音樂上載到Spotify等音樂串流平台，從而賺取佣金。

但憂的是，如果音樂可以被AI模型「創造」出來，那專業音樂人會否就此失業？從去年荷里活演員和劇作家對AI持續了5個月的抗議，不難看出這種憂慮已迫在眉睫。

事實上，這種憂慮也並非空穴來風。2017年Spotify上播放的音樂中，有87%是來自唱片公司的簽約歌手，但是到了2022年這個比率下降到了75%。截至2023年，AI已經生成了超過一億首樂曲，大概佔據了我們30%傾聽音樂的時間。業界人士預計，AI音樂的市場收益將會在2026年達到70億美元；到2030年，AI音樂將佔據50%的音樂市場份額。

數量抑或質素？

目前AI生成音樂的優勢在於速度和數量。美國初創公司Boomy聲稱，短短幾年AI生成樂曲已多達1800萬首；相比之下，Spotify上縱貫古今的曲目也只有一億首。可是，AI生成的音樂質素是否可以媲美專業音樂人的創作呢？目前AI的「創作」是基於過去的音樂，隨着數量的迅速增長，其質素會回歸平均（Regression to the mean）；而當大眾對新技術的新鮮感退潮後，對AI生成的音樂會否感到厭煩，轉而追捧音樂人的創作？又或者，AI生成音樂和音樂人的原創音樂是否可以「科學分工」，比如AI生成音樂可以作為成本較低的背景音樂，而音樂人的創作則在演唱會舞台上熠熠生輝。

人們對於數量和質素的要求，從來都不是捨此棄彼，如何優勢互補，值得AI公司和音樂人共同探索。

對AI生成音樂感到頭痛的，不僅僅有音樂人，更有靠音樂版權為生的唱片公司。AI模型之所以可以「創造」音樂，依賴的原材料就是現有唱片公司旗下的音樂。模型學習了這些音樂的特點，從而「創造」新的音樂。但是AI公司是否需要為這些原材料付版權費？AI創造出來的音樂，是否也應該受版權保護？

近年來，對這些問題的挑戰層出不窮。比如去年，環球唱片指控一家由亞馬遜和谷歌投資的AI公司Anthropic非法使用旗下的音樂，來訓練AI模型。Anthropic公司則聲稱使用現有的音樂訓練AI模型，不屬於盜版侵權。

挑戰還是機會？

由於AI技術的發展速度遠遠超過了現行知識產權法例的進展，這些無法迅速解決的問題，就成為了灰色地帶，為唱片公司和AI科技公司既帶來挑戰，也提供機會。回顧歷史，這不是唱片公司第一次遇到科技對版權的挑戰。早在二十世紀末，當音樂從CD變成MP3電子檔案，加上Napster等共享檔案平台的興起，盜版音樂泛濫讓不少唱片公司幾近關門大吉。唱片業花了10年時間，才開發出比傳統售賣CD更加有利可圖的商業模式，並最終與音樂串流平台在音樂版權上達成協議。

以史為鑑，隨着AI音樂勢不可擋，唱片公司不再視其為洪水猛獸，而是努力尋找新的商業模式，讓音樂版權在AI時代可帶來更大收益。華納音樂集團CEO Robert Kyncl曾表示：「我們不會簡單粗暴地拒絕阻止AI。」在推動音樂版權的法律界定和保護措施的同時，積極利用AI，一方面協助專業音樂人以更低成本、更快速度創作音樂，比如運用AI將播客（Podcasts）內容轉換成不同語言，讓不同國度的聽眾一飽耳福，甚至利用機器學習模型將披頭四主唱John Lennon在1973年留下的一首模糊音樂demo提取出來，不僅在50年後的今天「復活」了這首Now and Then，還讓披頭四的經典歌曲重新火了一把；另一方面，也訓練AI模型來精準定位侵權音樂，在必要時提出法律訴訟。更有甚者，使出「蘿蔔加大棒」的策略，一邊抗議AI公司侵權，一邊推進與AI公司合作，利用其版權優勢搶佔市場先機。

兩百年前，一曲《命運交響曲》讓我們感受到雙耳失聰的貝多芬對命運的抗爭；兩百年後的今天，如果貝多芬復活，AI是否可以化為他的耳朵，促發他的靈感，為我們帶來更多傳世佳作？兩百年前，貝多芬用音符吶喊出：「聽！命運在敲門！」兩百年後的今天，希望AI敲開的是人類創造力之門！

本欄逢周三刊登

范亭亭博士
港大經管學院市場學首席講師

Professor Heiwai Tang and Ms Shuyi Long

30 October 2024

In the 2024 Policy Address recently announced by the Chief Executive, mention is made of the Government’s plan to boost investment in the development of the digital economy, particularly in the digitalization of trade. Concrete measures include expediting the establishment of the Trade Single Window and forming a working group within the Hong Kong Monetary Authority to study the creation of a digital trade ecosystem with a focus on talent and infrastructure.

Soon after its release, the Policy Address sparked wide discussion in the community. Besides repeatedly writing about launching trade digitalization in Hong Kong time and again, we have also advocated for this issue publicly. Not only is digital trade an emerging trend in trade but it also presents a golden opportunity for the city’s growth. Hence, apart from the prerequisite hardware for building the Trade Single Window, it is crucial to concentrate on the transformation of trade models and to be fully prepared for the development of digital services and cross-border e-commerce.

Paperless trade

One key aspect of digital trade is paperless trade, which involves digitalizing and automating all procedures during the trade process. This entails converting paper texts into digital files and changing from manual to electronic vetting processes so as to reduce costs in labour, resources, and time. The Trade Single Window  in the Policy Address is tantamount to digital customs administration. This platform streamlines the processing of imports and exports by integrating all procedures, covering customs clearances, declarations, document completion and submission, as well as fee payments into a unified digital interface.

Not a novel concept that emerged in recent years, the Trade Single Window has long been introduced in various countries and has already been widely used. According to the 2023 United Nations Global Survey on Digital and Sustainable Trade Facilitation, over 40 countries worldwide have fully implemented the Trade Single Window System, including developed nations in Europe, the US, and Japan as well as developing countries such as Peru, Thailand, and Brazil. Having launched the first two phases of the Trade Single Window, the Hong Kong SAR Government plans to complete the final phase by 2026.

In addition to customs, international trade encompasses many other aspects ranging from shipping and goods collection to loans and insurance, plenty of which still require paper documents. A 2022 report of the World Trade Organization (WTO) estimates that each cross-border trade transaction involves at least 240 copies of 36 documents. Digitalizing the submission, vetting, and handling of all these documents will not only save paper and protect the environment but also streamline processes, saving much time and manpower. While establishing the Trade Single Window, Hong Kong should hasten the digitalization of other departments associated with trade, including updating regulations on the legal status of electronic documents and enhancing the digital infrastructure for processing them, thereby enabling Hong Kong to handle a larger trade volume.

Electronic commerce

Progress in paperless trade can be regarded as pivotal to digital trade and even the broader digital economy. However, digital trade does not merely change the flow of traditional goods trade to electronic mode. The WTO and the World Bank classify digital trade into digitally ordered trade and digitally deliverable trade. The former is characterized by e-commerce while the latter comprises the bulk of financial, legal, and consultation services. Both types of digital trade have enormous potential in the Hong Kong market. Although still in its infancy, e-commerce in the SAR has ample room for growth. Given that imports and exports of services are one of Hong Kong’s strengths, the digitalization of services trade is sure to usher in greater opportunities for this thriving sector.

E-commerce has become a massive market with a global income exceeding US$4 trillion and is expected to maintain rapid expansion for at least another decade. Hong Kong’s e-commerce market has also undergone dramatic development in recent years. Government statistics show that e-commerce sales were valued at over HK$30 billion in 2023, with clear signs of continued growth momentum ahead. Readers may be aware from their daily experiences that businesses like Hong Kong’s Yoho and HKTV Mall, the Mainland’s Taobao and Jingdong, and Amazon from overseas have become an increasingly important part of our lives, whether through their online platforms or retail sales.

Yet the Hong Kong’s e-commerce market still offers tremendous opportunities for development. At present, this sector accounts for approximately 8% of the SAR’s total retail sales—a percentage that pales in comparison to developed e-commerce markets such as Mainland China, the UK, and South Korea (each standing above 25%) but is also lower compared to neighbouring countries such as Japan, Taiwan, and Singapore. This relatively low market share implies that consumption potential remains largely unexploited. Meanwhile, numerous enterprises and merchants will have the opportunity to get a slice of the e-commerce pie.

The local e-commerce market is now in urgent need of improvement in the following two areas. First, e-merchant facilities, e.g. logistics and internet platforms, are not up to par. Despite the availability of next-day delivery, same-day delivery, and even delivery by the hour in the Mainland, Japan, and South Korea, it can still take up to three days for orders to be delivered from Kowloon to Hong Kong Island. Consumers encountering problems with their purchases still have to undergo complicated procedures, ranging from after-sales service and communications to returns and refunds. These inconveniences only offset the biggest advantage of online platforms―efficient shopping. Ever in pursuit of efficiency, Hongkongers may find it more convenient to shop by going out to local stores or travelling north to Shenzhen.

Second, for businesses looking to develop e-commerce capabilities, the lack of the right skills is another problem. With an insufficient talent pool in e-commerce, hiring is difficult even for big companies, let alone small and medium enterprises (SMEs). A report released by FedEx in 2022 reveals that 60% of Hong Kong’s SMEs find it difficult to hire personnel with e-commerce skills. E-commerce differs from conventional retailing in terms of management, sales, operations, and promotion. Hence, to many merchants, e-commerce professionals are a prerequisite for developing this business. The Hong Kong SAR Government should drum up support for talent training programmes at local higher education institutions and companies so as to quickly expand the talent reserve, thereby giving a boost to the e-commerce sector.

Trade in services

Services trade delivered through digital channels largely comprises professional services, including finance, law, education, healthcare, and information technology. In 2023, the world’s digital services exports amounted to over US$4 trillion, with the US, Mainland China, Japan, and India being the largest exporters.

Hong Kong being the world’s most services-oriented economy, the city’s services sector contributes to over 90% of its GDP, 60% of which is made up of services delivered through digital channels. Last year, the sector’s total production value exceeded HK$2.5 trillion, with the financial sector alone contributing more than HK$550 billion. With competitive advantages such as diverse services, a sound legal and judicial system, and an abundance of professionals, Hong Kong is second to none among the large services exporters mentioned above. Nevertheless, the digital services export figures simply do not do justice to these obvious advantages. In 2023, Hong Kong’s total services exports were valued at HK$700 billion, with digital services exports accounting for merely HK$300 billion (approximately US$45 billion). In comparison, Singapore’s digital services exports in the same year were valued at US$180 billion, nearly five times those of Hong Kong.

In the digital trade era, the challenges facing Hong Kong in leveraging the services sector’s distinct advantages can be attributed to the following reasons. First, digital services exports have been hindered by the incomplete progress in paperless trade. So long as required procedures of customs, banks, and the Government remain to be fully digitalized, digital services exports cannot be conducted on a large scale. In addition, unlike trade in goods, trade in services is also subject to various restrictions governing internet safety, cross-border data flow, and cross-border electronic payments. Since these issues cannot be resolved unilaterally, Hong Kong must negotiate with its trading partners to reach a consensus, making digital trade-related agreements indispensable. In this year’s Policy Address, the Government pledges to insert relevant provisions on digital trade and cross-border data flow into bilateral and multilateral trade agreements.

Last but not least, all sectors rely on platforms and opportunities to break into overseas markets, and the services sector is no exception. While sizeable companies in the services sector can probably explore export prospects on their own, SMEs are bound to stumble upon formidable challenges in following suit. They need the Government to provide them with information and connections, much like those offered for trade shows and exchange activities. The Government should consider taking similar measures to facilitate overseas visits and exchanges between companies in the services sector with foreign businesses. In addition, Government offices can be set up in Hong Kong’s major trading partner countries to help local enterprises to develop overseas markets. These initiatives would benefit the thriving services sector, propelling it to new heights.

Dr Tingting Fan

4 December 2024

When we listen to music through our earphones these days, do we realize that over 30% of it is already produced by AI? Last year, the AI-generated song “Heart on My Sleeve” got 20 million hits on Spotify. Early this year, the burgeoning AI-generated music scene drove Universal Music, the world’s leading music company, to remove all its records from TikTok, the largest platform for short-form mobile videos. After the two companies reached an agreement in mid-2024, TikTok agrees to label all AI music footage accordingly. The impact of AI technology on the music industry has been fast and furious.

Cause for celebration or concern

With the progress in AI technology, major technology companies have extended their reach over the music industry. For example, the text-to-music model named MusicGen was launched by Meta to users in 2023. The Stable Audio 2.0 model, introduced by Stability AI this year, even allows users to upload existing music to generate new tracks in a completely different style. The acoustic quality is comparable to that of a vinyl record.

It is a cause for celebration because AI enables ordinary people, who are not music professionals, to not only “create” music with ease but also earn money from these “creative” works. Boomy, an American start-up, supports users in uploading their AI-generated music to Spotify and other streaming platforms for a commission.

That being said, it is a cause for concern because if music can be “created” by an AI model, would professional musicians find themselves out of a job? Given the five consecutive months of protest from Hollywood actors and screenwriters in 2023, the looming fear is clear as day.

As a matter of fact, the great concern is not unjustified. In 2017, 87% of music tracks played on Spotify were from singers signed with record labels. By 2022, this percentage fell to 75%. As of 2023, over 100 million pieces of music were generated by AI, taking up around 30% of our music-listening time. The revenue generated by the AI music market is projected by industry members to reach US$7 billion by 2026, while AI music is expected to have a 50% share of the music sector by 2030.

Quantity or quality

At present, the advantages of AI-generated music lie in speed and quantity. Boomy claims that in just a few years, there are already 18 million pieces of AI-generated music, whereas only 100 million pieces of old and new music spanning all time periods have found their way to Spotify so far. Nevertheless, does the quality of AI-composed music rival that of the creative works by professional musicians? So far, the works created by AI have been based on past music. With the rapid growth in quantity, the quality of AI-generated music will eventually regress to the mean. When the excitement over this new technology wears off for the public, will people tire of AI-produced music and turn to works by music artists? Alternatively, one wonders if a “scientific division of labour” is possible, whereby AI-generated music serves as low-cost background music while the concert stage is reserved for music artists’ works to shine.

When it comes to people’s requirements for music, quantity and quality are never mutually exclusive. Striking a balance between the two is something that both AI companies and music artists should explore.

As a matter of fact, both music artists and record companies, which rely on music copyrights to survive, are on the receiving end of AI-generated music’s vexing challenge. The music works owned by record companies provide the raw materials for AI models to “create” new music after learning their characteristics. But should AI companies be obliged to pay royalties for these raw materials? And should AI-created music be under copyright protection?

Recent years have seen increasing challenges arising from these issues. In 2023, for instance, Universal Music accused Anthropic, an AI company with investments from Google and Amazon of illegally using works owned by Universal Music to train Anthropic’s AI models. In its defence, the company claims that using existing music to train AI models does not constitute copyright infringement.

Challenge or opportunity

Since advancements in AI technology have far outpaced developments in intellectual-property laws, these problems without quick solutions have become a grey area, presenting both challenges and opportunities for record companies and AI technology companies. In retrospect, this is not the first time music publishers have encountered copyright challenges. The late 20th century saw the migration of music from CDs to electronic MP3 files, which, coupled with the rise of sharing platforms like Napster, led to rampant pirated music that pushed many record companies out of business. It took an entire decade for record companies to develop a business model more profitable than the traditional approach of selling CDs and to eventually reach agreements with music streaming platforms regarding music copyright.

Taking lessons from history and embracing the unstoppable trend of AI music, record companies no longer regard it as an uncontrollable beast. Instead, they are striving to devise a new business model that can enable music copyrights to bring greater profit in the AI era. Robert Kyncl, CEO of Warner Music Group, once says that simply rejecting AI and fighting against it is out of the question. In promoting legal definition and protection of music copyright, record companies actively use AI to facilitate music creation by professional artists in cheaper and faster ways on the one hand. For example, AI is harnessed to produce multilingual versions of podcasts for the enjoyment of audiences around the world. Machine learning has even been used to extract a muddled demo song left behind by the Beatles’ lead singer, John Lennon, in 1973. The recent release not only gave new life to the song “Now and Then” but also reignited enthusiasm for their Beatles’ classic ballads. On the other hand, AI models are also trained to precisely detect copyright-infringing music for litigation purposes, if necessary. Additionally, record companies may even roll out a two-pronged carrot-and-stick strategy to protest against copyright infringement by AI companies while leveraging their advantage as copyright owners to become market pioneers through closer cooperation with AI companies.

Two centuries ago, the Fate Symphony strikes a chord with us, helping us to empathize with Beethoven’s struggle against destiny after losing his hearing. Two centuries later, if Beethoven came back to life, could AI restore his hearing to inspire him to compose even more masterpieces for posterity? Two hundred years ago, through musical notes, Beethoven issued the rallying cry: “Listen! Fate is knocking at the door!” Today, two hundred years later, hopefully the door opened by AI will lead to a new era of human creativity!

美國大選結果折射出的避責文化

互相指摘或卸責，也許只是個人沒擔當的怯懦行為，但放大到社會層面，就足以產生混淆公眾視聽的惡果。政治人物往往透過彼此指摘來轉移視線，力求貶低對手而抬高自己。在競爭白熱化的選舉中，不惜一切推卸責任已成政客的慣技，或對選舉結果以至未來管治和政策帶來難以想像的衝擊。

民主黨敗選背後

本月美國總統選舉結果塵埃落定，共和黨特朗普以壓倒性姿態勝出。賀錦麗慘敗後，其所屬民主黨內隨即出現大舉卸責現象，矛頭直指拜登，歸咎他未能及時退選，陷賀錦麗於尷尬境地。不少黨內成員亦認為拜登年老退化，不受選民歡迎，雖然他及後宣布退選，賀錦麗仍因受選民支持度不足，未能於明年入主白宮。

與此同時，民主黨在國會改選中失去參議院和眾議院的控制權，較4年前表現更糟。根據《紐約時報》的分析，自拜登在2020年出任總統以來，美國3100多個縣的選民大都轉向右傾。民主黨向來標榜的支持墮胎權和民主立場，無法像經濟和移民等迫切議題引起選民共鳴。

儘管美國失業率現正維持在歷史低位，股市暢旺，但物價高、房租貴也是事實。拜登任內，物價上漲超過20%。康奈爾大學的經濟學家巴蘇（Kaushik Basu）指出，各種經濟指標之中，通脹對政治影響最大。一般人無需數據，也對通脹有切身感受。再者，《金融時報》的分析顯示，在今年舉行選舉的10個國家中，執政黨的表現都不如上屆選舉，相信也與高通脹有關。

根據民調，三分之二的美國選民對經濟給予劣評，收入較低的一群傾向於支持特朗普。2020年，他以15個百分點的差距失去收入介乎5萬至10萬的選民，但在這次選舉中卻逆轉獲勝。民主黨人似乎忽略了馬斯洛的需求層次理論（Maslow’s Hierarchy of Needs）：基本需要（如財務穩健和身心健康）必須先行，然後再滿足其他方面。在競選活動中，民主黨聚焦於民主等議題而忽略經濟。曾經是該黨核心的工人階級選民不再予以支持，因愈來愈多人按自身的經濟利益來投票。黨內對敗選結果莫衷一是，更出現互相指摘。如此反應，是否就能把選票贏回來？答案不言而喻。政治指摘伎倆層出不窮，皆因政黨或領導人藉此進行政治操弄，以便大權在握。

企業卸責文化

在商業環境中，互相指摘確也頗為普遍。譬如一家公司面臨存亡危機，責任的分配將直接影響其股價和投資者的信心。假使管理層只管找替罪羊，哪怕是象徵式的代罪羔羊，公司亦難逃衰敗的厄運。從管理學的研究可見，將公司失敗歸咎於外部因素的管理層，其整體表現往往不及承認自身責任並自我反省的公司。在瀕臨破產的企業中，可以看到不少經理將業績欠佳委過於其他部門。相反，管理層若有責任感，則有可能轉虧為盈，讓業務重上軌道。前事不忘，後事之師，只有汲取教訓，才能避免重蹈覆轍。

此外，互相指摘也足以助長風險規避文化，員工因害怕受責備而不敢主動行事，或礙於不願分享想法而窒礙創意。眾所周知，成功的企業有賴暢順的運作；管理層必須致力培養團隊合作精神，以解決公司內部的分歧。

「無過失」調查的啟示

反觀一些行業早已認識到指摘的弊端，例如航空業所以在降低意外事故一環取得成效，很大程度上受惠於「無過失」調查的程序。在美國，負責調查有關事故的國家運輸安全委員會明確表示，調查目的並非追究責任，而是找出問題並提出建議，以防同類事件重演。航空業不進行追責的事後調查，為現代航空安全奠下重要基石。

這種調查方式有助於建立開放的安全文化，鼓勵業界報憂，最終目的是確保減少意外事故。英國的航空監管機構在誠實錯誤和其他錯誤之間劃界線，也是個好的起點。航空公司致力於營造一種文化，使機師不會因為與其經驗和培訓相符的決定或疏忽而受到懲罰。這種做法並非完全免責，只是將責任範圍收窄而已。

醫療保健領域也面臨類似情況。一旦發生醫療事故，世界各地對病人的補償制度各有不同。例如英國依賴找出過失的訴訟程序，而紐西蘭則是全球最早實施醫療事故處理制度的國家。紐西蘭率先以「無過失補償」的程序來處理醫療事故，並於1974年成立意外補償局負責，接受因工作、交通或醫療事故導致的傷害賠償申請。在這一制度下，無論醫療措施或副作用造成的傷害是否可以避免，病人均可向補償局提出申請。只要問題與醫療診斷或決策相關，申請便可獲批准。該制度推行後，除非醫療人員的行為嚴重違法，否則紐西蘭患者幾乎無法向醫療機構提出訴訟。

在航空和醫療領域，從錯誤中學習的動機特強，因為從業員在工作中生命隨時受到威脅，安全無疑至關重要。因此，軟件工程師和開發人員經常進行「無過失的事後分析」，以調查網站失靈或伺服器故障等問題。一般人不易理解這種不追責的思維，心理學家James Reason在1990年代為此提出一個框架，以釋除大眾對無能和犯錯者逃避責罰的疑慮。

問責而非卸責

要逃避指摘其實並非易事。一、當事人為了避責往往要大費心力，但指摘別人反而是毫不費力的快速反應，而且容易令人入信。至於記錄錯誤並確保流程得以改進，則難免涉及結構性的變化。例如無過失事後分析長期以來已屬谷歌企業文化的一部分，該公司為此提供模板、反饋和討論小組。二、企業管理層既然大權在握，指摘屬下僱員也就輕而易舉。

加州大學聖地牙哥分校和新加坡南洋理工大學的學者最近合作發表一篇研究論文，指出當權者往往認為其他人會將失敗歸咎於他們。在一項實驗中，參與者被隨機分配為主管或工人，然後檢視有關錯誤的紀錄。參與者都收到道歉信，聲明網絡連接不穩定，以致任務無法正常完成。結果扮演主管者每多認定抄寫員應為失誤負責，主張剋扣其報酬。由此可見掌權與施罰之間的因果關係。

指摘別人似乎也具傳染性。2009年，心理學學者David Sherman 和 John Klein發表合著論文【註】，其中一個實驗要求參與者閱讀有關政治失敗的新聞，然後寫下政客的過失。讀到關於政客將失敗歸咎於特殊利益的報道時，參與者更可能將自己的失敗責任推卸給別人。至於讀到政客承擔責任的參與者，則更可能肯為自身的不足負責。同理，管理高層若輕易指摘別人，公司員工也會有樣學樣。如此一來，不難衍生出一種推卸責任的指摘文化。

不同文化對於失責和指摘的容忍度不盡相同。例如集體主義可能導致共同指摘，而在個人主義的文化中，個人指摘則較常見。相互指摘的經濟學強調人類行為與經濟結果之間的相互作用。了解這些動態關係當有助於機構創造出更具建設性的環境，減少諉過於人，以鼓勵問責和合作。

註：Sherman, D. K. and John M. Klein, “Failure to Blame: The Effect of Collective Blame on Self-Attribution.” Psychological Science, 2009.

謝國生博士
港大經管學院金融學首席講師、新界鄉議局當然執行委員

何敏淙先生
香港大學附屬學院講師

（本文同時於二零二四年十一月二十七日載於《信報》「龍虎山下」專欄）

美國大選結果折射出的避責文化

互相指摘或卸責，也許只是個人沒擔當的怯懦行為，但放大到社會層面，就足以產生混淆公眾視聽的惡果。政治人物往往透過彼此指摘來轉移視線，力求貶低對手而抬高自己。在競爭白熱化的選舉中，不惜一切推卸責任已成政客的慣技，或對選舉結果以至未來管治和政策帶來難以想像的衝擊。

民主黨敗選背後

本月美國總統選舉結果塵埃落定，共和黨特朗普以壓倒性姿態勝出。賀錦麗慘敗後，其所屬民主黨內隨即出現大舉卸責現象，矛頭直指拜登，歸咎他未能及時退選，陷賀錦麗於尷尬境地。不少黨內成員亦認為拜登年老退化，不受選民歡迎，雖然他及後宣布退選，賀錦麗仍因受選民支持度不足，未能於明年入主白宮。

與此同時，民主黨在國會改選中失去參議院和眾議院的控制權，較4年前表現更糟。根據《紐約時報》的分析，自拜登在2020年出任總統以來，美國3100多個縣的選民大都轉向右傾。民主黨向來標榜的支持墮胎權和民主立場，無法像經濟和移民等迫切議題引起選民共鳴。

儘管美國失業率現正維持在歷史低位，股市暢旺，但物價高、房租貴也是事實。拜登任內，物價上漲超過20%。康奈爾大學的經濟學家巴蘇（Kaushik Basu）指出，各種經濟指標之中，通脹對政治影響最大。一般人無需數據，也對通脹有切身感受。再者，《金融時報》的分析顯示，在今年舉行選舉的10個國家中，執政黨的表現都不如上屆選舉，相信也與高通脹有關。

根據民調，三分之二的美國選民對經濟給予劣評，收入較低的一群傾向於支持特朗普。2020年，他以15個百分點的差距失去收入介乎5萬至10萬的選民，但在這次選舉中卻逆轉獲勝。民主黨人似乎忽略了馬斯洛的需求層次理論（Maslow’s Hierarchy of Needs）：基本需要（如財務穩健和身心健康）必須先行，然後再滿足其他方面。在競選活動中，民主黨聚焦於民主等議題而忽略經濟。曾經是該黨核心的工人階級選民不再予以支持，因愈來愈多人按自身的經濟利益來投票。黨內對敗選結果莫衷一是，更出現互相指摘。如此反應，是否就能把選票贏回來？答案不言而喻。政治指摘伎倆層出不窮，皆因政黨或領導人藉此進行政治操弄，以便大權在握。

企業卸責文化

在商業環境中，互相指摘確也頗為普遍。譬如一家公司面臨存亡危機，責任的分配將直接影響其股價和投資者的信心。假使管理層只管找替罪羊，哪怕是象徵式的代罪羔羊，公司亦難逃衰敗的厄運。從管理學的研究可見，將公司失敗歸咎於外部因素的管理層，其整體表現往往不及承認自身責任並自我反省的公司。在瀕臨破產的企業中，可以看到不少經理將業績欠佳委過於其他部門。相反，管理層若有責任感，則有可能轉虧為盈，讓業務重上軌道。前事不忘，後事之師，只有汲取教訓，才能避免重蹈覆轍。

此外，互相指摘也足以助長風險規避文化，員工因害怕受責備而不敢主動行事，或礙於不願分享想法而窒礙創意。眾所周知，成功的企業有賴暢順的運作；管理層必須致力培養團隊合作精神，以解決公司內部的分歧。

「無過失」調查的啟示

反觀一些行業早已認識到指摘的弊端，例如航空業所以在降低意外事故一環取得成效，很大程度上受惠於「無過失」調查的程序。在美國，負責調查有關事故的國家運輸安全委員會明確表示，調查目的並非追究責任，而是找出問題並提出建議，以防同類事件重演。航空業不進行追責的事後調查，為現代航空安全奠下重要基石。

這種調查方式有助於建立開放的安全文化，鼓勵業界報憂，最終目的是確保減少意外事故。英國的航空監管機構在誠實錯誤和其他錯誤之間劃界線，也是個好的起點。航空公司致力於營造一種文化，使機師不會因為與其經驗和培訓相符的決定或疏忽而受到懲罰。這種做法並非完全免責，只是將責任範圍收窄而已。

醫療保健領域也面臨類似情況。一旦發生醫療事故，世界各地對病人的補償制度各有不同。例如英國依賴找出過失的訴訟程序，而紐西蘭則是全球最早實施醫療事故處理制度的國家。紐西蘭率先以「無過失補償」的程序來處理醫療事故，並於1974年成立意外補償局負責，接受因工作、交通或醫療事故導致的傷害賠償申請。在這一制度下，無論醫療措施或副作用造成的傷害是否可以避免，病人均可向補償局提出申請。只要問題與醫療診斷或決策相關，申請便可獲批准。該制度推行後，除非醫療人員的行為嚴重違法，否則紐西蘭患者幾乎無法向醫療機構提出訴訟。

在航空和醫療領域，從錯誤中學習的動機特強，因為從業員在工作中生命隨時受到威脅，安全無疑至關重要。因此，軟件工程師和開發人員經常進行「無過失的事後分析」，以調查網站失靈或伺服器故障等問題。一般人不易理解這種不追責的思維，心理學家James Reason在1990年代為此提出一個框架，以釋除大眾對無能和犯錯者逃避責罰的疑慮。

問責而非卸責

要逃避指摘其實並非易事。一、當事人為了避責往往要大費心力，但指摘別人反而是毫不費力的快速反應，而且容易令人入信。至於記錄錯誤並確保流程得以改進，則難免涉及結構性的變化。例如無過失事後分析長期以來已屬谷歌企業文化的一部分，該公司為此提供模板、反饋和討論小組。二、企業管理層既然大權在握，指摘屬下僱員也就輕而易舉。

加州大學聖地牙哥分校和新加坡南洋理工大學的學者最近合作發表一篇研究論文，指出當權者往往認為其他人會將失敗歸咎於他們。在一項實驗中，參與者被隨機分配為主管或工人，然後檢視有關錯誤的紀錄。參與者都收到道歉信，聲明網絡連接不穩定，以致任務無法正常完成。結果扮演主管者每多認定抄寫員應為失誤負責，主張剋扣其報酬。由此可見掌權與施罰之間的因果關係。

指摘別人似乎也具傳染性。2009年，心理學學者David Sherman 和 John Klein發表合著論文【註】，其中一個實驗要求參與者閱讀有關政治失敗的新聞，然後寫下政客的過失。讀到關於政客將失敗歸咎於特殊利益的報道時，參與者更可能將自己的失敗責任推卸給別人。至於讀到政客承擔責任的參與者，則更可能肯為自身的不足負責。同理，管理高層若輕易指摘別人，公司員工也會有樣學樣。如此一來，不難衍生出一種推卸責任的指摘文化。

不同文化對於失責和指摘的容忍度不盡相同。例如集體主義可能導致共同指摘，而在個人主義的文化中，個人指摘則較常見。相互指摘的經濟學強調人類行為與經濟結果之間的相互作用。了解這些動態關係當有助於機構創造出更具建設性的環境，減少諉過於人，以鼓勵問責和合作。

註：Sherman, D. K. and John M. Klein, “Failure to Blame: The Effect of Collective Blame on Self-Attribution.” Psychological Science, 2009.

謝國生博士
港大經管學院金融學首席講師、新界鄉議局當然執行委員

何敏淙先生
香港大學附屬學院講師

（本文同時於二零二四年十一月二十七日載於《信報》「龍虎山下」專欄）

特朗普會辭退鮑威爾嗎？

特朗普在成功當選下屆美國總統後，迅即籌組內閣。從人選來看，差不多每個任命都是要顛覆原有建制，這為未來數年的美國及國際社會增添濃厚不確定性。明年1月，美國各政府部門都有新主管上任，然而有一個重要職位不會因總統換屆而改變人選，那便是聯儲局主席。現任主席鮑威爾由特朗普提名和參議院確認，但隨後因加息引起特朗普的不滿。看來若法律或政治成本低的話，特朗普也會以親信取代鮑威爾，並將聯儲局來個翻天覆地的改動。

聯儲局與貨幣政策的重要性不言而喻，在這次總統選舉中更清晰可見。不少調查都顯示經濟是選民最關心議題，而40年來首次出現的高通脹，正是民主黨失敗的一個主要原因。這次通脹在2021年中已明顯呈現，但聯儲局在翌年3月才首次加息，未能先發制人，通脹因而冒升至2022年6月的9.1%。雖然通脹在今年中已放緩至約3%，惟物價水平仍高企。由2021年4月至本年10月，消費者物價指數上升了18.3%，這都被選民算在拜登和賀錦麗頭上，同期間工資雖然有增長，卻被看為個人努力的成果。

面對通脹，拜登政府也有一些應對措施，例如出售40%的戰略石油儲備以紓緩能源價格，但畢竟不若貨幣政策之有效。同時，民主黨被傳統思維限制，在經濟議題上一貫重就業輕通脹，看見失業率徘徊於歷史最低水平的4%以下，便認為已贏得民心，忽略了40年來通脹都處於低水平，約50歲以下選民都是首次面對無端失去大幅購買力的困境，而把通脹歸咎於企業提高價格以謀取暴利的論述，實際上並沒有解決問題。聯儲局這次對通脹反應過慢，加上其他政治經濟因素，使一些政客產生把貨幣政策收歸政府行政部門的意圖。諷刺的是，特朗普這次勝出的一個原因是兩年多的通貨膨脹，但他二進白宮後迅速推行的高關稅，將大幅提高美國物價。

美國大選結果揭曉後不久，剛好是聯儲局議息會議結束，鮑威爾在記者會上被問及，如果特朗普要他辭職會否接受，他只簡單回應一個「不」字，然後是好幾秒鐘的冷場，無聲勝有聲地道出他與特朗普之間的張力。特朗普2017年以鮑威爾取代耶倫任聯儲局主席；其後聯儲局在2017及2018兩年加息7次，每次0.25厘，把2008年金融海嘯後接近零的利率提高至較正常水平，但特朗普認為提高利率會影響他任內經濟，公開稱鮑威爾是敵人，他和他的團隊是笨蛋（bonehead）等。這次競選期間，特朗普數次提過總統要有權決定利率，認為自己賺錢很多，也很成功，在很多情況下比聯儲局決策人和聯儲局主席有更佳直覺。其後他又自辯，說不是要直接控制利率，只是認為總統可如其他人一樣就利率政策表達意見。他警告在競選期間減息有助民主黨競選，但又多次提及當選後會減息，儘管在目前框架下，他沒有這個權力。

鮑仍可掌FOMC 誕兩權力中心

聯儲局成立的法律依據，是1913年國會通過的《聯邦儲備法案》（Federal Reserve Act）。聯儲局有3個主要組成部分，分別是聯邦儲備理事會（Federal Reserve Board of Governors）、12家聯邦儲備銀行（Federal Reserve Banks）和聯邦公開市場委員會（Federal Open Market Committee，簡稱FOMC）。嚴格來說，《聯邦儲備法案》中沒有聯儲局主席一職，只有聯邦儲備理事會主席。理事會有7位成員，都由總統提名和參議院確認，任期14年，原則上不能續任，而且任期平均分布，大約每兩年委任一名新成員。此外，理事會主席及兩位副主席也是由總統在7人中提名及參議院確認，每個任期4年。現時鮑威爾作為理事會主席任期到2026年5月，但作為理事會成員，他的任期到2028年1月才結束。

至於那12家聯邦儲備銀行，分別對全國12個區域提供服務，主席則由該銀行的董事物色及委任，不在總統權力之內。聯儲局的貨幣政策由FOMC制定，FOMC設有12位成員，包括理事會的7位成員及紐約聯邦儲備銀行主席，其餘4位則由餘下11家聯邦儲備銀行的主席輪流擔任。其他聯邦儲備銀行主席可以出席FOMC會議，但沒有投票權。《聯邦儲備法案》沒有說明誰是FOMC主席，傳統上，FOMC的12位成員都會推舉理事會主席為FOMC主席及紐約聯邦儲備銀行主席為FOMC副主席。也就是說，FOMC的正、副主席並非由總統提名委任。

現時鮑威爾在聯儲局中有幾個職位，分別是理事會的主席和成員，以及FOMC的主席和成員。特朗普要辭退鮑威爾，須先從他理事會成員職位着手，但要解除理事會成員職務，必須要有一個合適和有力的理由，如瀆職或玩忽職守等。若以政策看法與總統或白宮相左為理由，自然難以服眾。如果只解除鮑威爾理事會主席職務，仍然留他為理事會成員，程序或許較為簡單或阻力較小，不過他仍可被推舉為FOMC主席，和今天一樣在每次議息會議後面對傳媒。若總統另外提名理事會主席，會使聯儲局內出現兩個權力中心。無論是解僱的過程或結果，都會帶來金融市場震盪。

和這點有關的是，目前可能出現的一個政策矛盾。FOMC是合法制定貨幣政策的單位，如上所述，它包括了7位理事會成員和5位聯邦儲備銀行主席。前者由總統提名委任，後者的委任和總統無關。與此同時，法律又把銀行在聯儲局儲備金的利息決定權授予理事會。在一般情況下，銀行儲備的利息與FOMC制定的聯邦基金利率步伐一致、相輔相成。但假如FOMC認為應該提高聯邦基金利率，較容易受總統影響的理事會卻降低銀行儲備金利率，政策矛盾便出現。

特朗普若不能或不想付出太高成本辭退鮑威爾，那他仍可以等到有空缺時委任志同道合者為理事會成員或主席，去影響聯儲局的貨幣政策，不過這有一定的困難。他在第一任總統期內提名了3人，均過不了參議院，其中女經濟學者謝爾頓（Judy Shelton）因主張美元與黃金掛鈎及對聯儲局獨立性置疑而失去一些共和黨參議員的支持。

有趣的是，被特朗普考慮做財政部長的貝桑（Scott Bessent），建議特朗普上任後即提名及爭取參議院提早確認新的理事會主席人選，作為影子主席，架空鮑威爾。影子主席可就貨幣政策發言，在慣常的前瞻指引（forward guidance）做法下，市場會比較聽取影子主席意見而不理會任期只到2026年的鮑威爾。貝桑說這只是他個人意見，而非特朗普的意見，但又說曾和特朗普討論，並有把這個想法與特朗普的顧問分享。未知貝桑這一招會否在未來美國黨爭中被重複使用，導致政壇幻影重重？

陸炎輝博士
港大經管學院榮譽副教授

（本文同時於二零二四年十一月二十日載於《信報》「龍虎山下」專欄）

特朗普會辭退鮑威爾嗎？

特朗普在成功當選下屆美國總統後，迅即籌組內閣。從人選來看，差不多每個任命都是要顛覆原有建制，這為未來數年的美國及國際社會增添濃厚不確定性。明年1月，美國各政府部門都有新主管上任，然而有一個重要職位不會因總統換屆而改變人選，那便是聯儲局主席。現任主席鮑威爾由特朗普提名和參議院確認，但隨後因加息引起特朗普的不滿。看來若法律或政治成本低的話，特朗普也會以親信取代鮑威爾，並將聯儲局來個翻天覆地的改動。

聯儲局與貨幣政策的重要性不言而喻，在這次總統選舉中更清晰可見。不少調查都顯示經濟是選民最關心議題，而40年來首次出現的高通脹，正是民主黨失敗的一個主要原因。這次通脹在2021年中已明顯呈現，但聯儲局在翌年3月才首次加息，未能先發制人，通脹因而冒升至2022年6月的9.1%。雖然通脹在今年中已放緩至約3%，惟物價水平仍高企。由2021年4月至本年10月，消費者物價指數上升了18.3%，這都被選民算在拜登和賀錦麗頭上，同期間工資雖然有增長，卻被看為個人努力的成果。

面對通脹，拜登政府也有一些應對措施，例如出售40%的戰略石油儲備以紓緩能源價格，但畢竟不若貨幣政策之有效。同時，民主黨被傳統思維限制，在經濟議題上一貫重就業輕通脹，看見失業率徘徊於歷史最低水平的4%以下，便認為已贏得民心，忽略了40年來通脹都處於低水平，約50歲以下選民都是首次面對無端失去大幅購買力的困境，而把通脹歸咎於企業提高價格以謀取暴利的論述，實際上並沒有解決問題。聯儲局這次對通脹反應過慢，加上其他政治經濟因素，使一些政客產生把貨幣政策收歸政府行政部門的意圖。諷刺的是，特朗普這次勝出的一個原因是兩年多的通貨膨脹，但他二進白宮後迅速推行的高關稅，將大幅提高美國物價。

美國大選結果揭曉後不久，剛好是聯儲局議息會議結束，鮑威爾在記者會上被問及，如果特朗普要他辭職會否接受，他只簡單回應一個「不」字，然後是好幾秒鐘的冷場，無聲勝有聲地道出他與特朗普之間的張力。特朗普2017年以鮑威爾取代耶倫任聯儲局主席；其後聯儲局在2017及2018兩年加息7次，每次0.25厘，把2008年金融海嘯後接近零的利率提高至較正常水平，但特朗普認為提高利率會影響他任內經濟，公開稱鮑威爾是敵人，他和他的團隊是笨蛋（bonehead）等。這次競選期間，特朗普數次提過總統要有權決定利率，認為自己賺錢很多，也很成功，在很多情況下比聯儲局決策人和聯儲局主席有更佳直覺。其後他又自辯，說不是要直接控制利率，只是認為總統可如其他人一樣就利率政策表達意見。他警告在競選期間減息有助民主黨競選，但又多次提及當選後會減息，儘管在目前框架下，他沒有這個權力。

鮑仍可掌FOMC 誕兩權力中心

聯儲局成立的法律依據，是1913年國會通過的《聯邦儲備法案》（Federal Reserve Act）。聯儲局有3個主要組成部分，分別是聯邦儲備理事會（Federal Reserve Board of Governors）、12家聯邦儲備銀行（Federal Reserve Banks）和聯邦公開市場委員會（Federal Open Market Committee，簡稱FOMC）。嚴格來說，《聯邦儲備法案》中沒有聯儲局主席一職，只有聯邦儲備理事會主席。理事會有7位成員，都由總統提名和參議院確認，任期14年，原則上不能續任，而且任期平均分布，大約每兩年委任一名新成員。此外，理事會主席及兩位副主席也是由總統在7人中提名及參議院確認，每個任期4年。現時鮑威爾作為理事會主席任期到2026年5月，但作為理事會成員，他的任期到2028年1月才結束。

至於那12家聯邦儲備銀行，分別對全國12個區域提供服務，主席則由該銀行的董事物色及委任，不在總統權力之內。聯儲局的貨幣政策由FOMC制定，FOMC設有12位成員，包括理事會的7位成員及紐約聯邦儲備銀行主席，其餘4位則由餘下11家聯邦儲備銀行的主席輪流擔任。其他聯邦儲備銀行主席可以出席FOMC會議，但沒有投票權。《聯邦儲備法案》沒有說明誰是FOMC主席，傳統上，FOMC的12位成員都會推舉理事會主席為FOMC主席及紐約聯邦儲備銀行主席為FOMC副主席。也就是說，FOMC的正、副主席並非由總統提名委任。

現時鮑威爾在聯儲局中有幾個職位，分別是理事會的主席和成員，以及FOMC的主席和成員。特朗普要辭退鮑威爾，須先從他理事會成員職位着手，但要解除理事會成員職務，必須要有一個合適和有力的理由，如瀆職或玩忽職守等。若以政策看法與總統或白宮相左為理由，自然難以服眾。如果只解除鮑威爾理事會主席職務，仍然留他為理事會成員，程序或許較為簡單或阻力較小，不過他仍可被推舉為FOMC主席，和今天一樣在每次議息會議後面對傳媒。若總統另外提名理事會主席，會使聯儲局內出現兩個權力中心。無論是解僱的過程或結果，都會帶來金融市場震盪。

和這點有關的是，目前可能出現的一個政策矛盾。FOMC是合法制定貨幣政策的單位，如上所述，它包括了7位理事會成員和5位聯邦儲備銀行主席。前者由總統提名委任，後者的委任和總統無關。與此同時，法律又把銀行在聯儲局儲備金的利息決定權授予理事會。在一般情況下，銀行儲備的利息與FOMC制定的聯邦基金利率步伐一致、相輔相成。但假如FOMC認為應該提高聯邦基金利率，較容易受總統影響的理事會卻降低銀行儲備金利率，政策矛盾便出現。

特朗普若不能或不想付出太高成本辭退鮑威爾，那他仍可以等到有空缺時委任志同道合者為理事會成員或主席，去影響聯儲局的貨幣政策，不過這有一定的困難。他在第一任總統期內提名了3人，均過不了參議院，其中女經濟學者謝爾頓（Judy Shelton）因主張美元與黃金掛鈎及對聯儲局獨立性置疑而失去一些共和黨參議員的支持。

有趣的是，被特朗普考慮做財政部長的貝桑（Scott Bessent），建議特朗普上任後即提名及爭取參議院提早確認新的理事會主席人選，作為影子主席，架空鮑威爾。影子主席可就貨幣政策發言，在慣常的前瞻指引（forward guidance）做法下，市場會比較聽取影子主席意見而不理會任期只到2026年的鮑威爾。貝桑說這只是他個人意見，而非特朗普的意見，但又說曾和特朗普討論，並有把這個想法與特朗普的顧問分享。未知貝桑這一招會否在未來美國黨爭中被重複使用，導致政壇幻影重重？

陸炎輝博士
港大經管學院榮譽副教授

（本文同時於二零二四年十一月二十日載於《信報》「龍虎山下」專欄）

Today, in the face of a complex set of Environmental, Social, and Governance (ESG) indicators and their increasingly diverse practices, ESG is no longer a marginal issue in the corporate world. How to enhance ESG competitiveness through precise policy implementation and high-efficiency management has become a burning question for enterprises. With relatively limited resources at their disposal, what ESG strategies should companies prioritize?

Optimizing cost-effectiveness with ESG ratings in mind

When mapping ESG strategies in the past, companies tended to adopt the indicators and weightings of ESG rating agencies as the basis in order to achieve higher ratings. Such an approach is rational as it aligns with the rating standards investors rely upon, thus enabling enterprises to gain the upper hand in compliance, financing, etc.

However, the burgeoning ESG rating agencies and their diversified rating standards in recent years have brought new challenges to businesses. Incomplete statistics show that there are now around 600 rating bodies worldwide, each selecting different indicators and weightings. Existing research indicates that the correlation among ratings from different agencies is weak (see Table), particularly in terms of Social and Governance themes.

Table   Correlation coefficients among ESG ratings of different agencies

Note: This table compares the correlation coefficients of ESG ratings of several mainstream agencies. SA, SP, MO, RE, KL, and MS represent Sustainalytics, S&P Global, Moody’s ESG, Refinitiv, KLD, and MSCI respectively. For example, the first column in the Table indicates that the ESG correlation coefficient between KLD and Sustainalytics is 0.53; the correlation coefficient for the E theme is 0.59, the correlation coefficient for the S theme is 0.31; and the correlation coefficient for the G theme is 0.02. These research results are derived from Berg et al. (2022).

Source: Florian Berg, Julian F Kölbel, Roberto Rigobon. “Aggregate Confusion: The Divergence of ESG Ratings.” Review of Finance, Vol 26, Issue 6 (2022): 1315–1344.

The uncertainty of such a rating approach produces several problems. First, even businesses that have invested heavily in ESG can still receive low ratings from some rating agencies. Second, companies that use ESG as a means of greenwashing or window dressing can instead get high ratings from some agencies. Coupled with the fact that many agencies keep churning out all sorts of league tables and awards for profit, the credibility of ESG ratings is going south. The resulting uncertainty over decisions based on ESG ratings poses formidable challenges for a wide range of decision-makers, including enterprises and investors.

To address this problem, we believe that, on the one hand, it is necessary to regulate the ESG-rating market to promote greater transparency of rating methodologies. On the other hand, enterprises should also further assess the actual costs and benefits of each ESG action and initiative so as to facilitate more rational ESG practices.

Specifically, enterprises should identify ESG actions conducive to not only social benefits but also effective cost control. By accurately identifying and prioritizing the implementation of these ESG measures, companies can ensure better value for money for each and every input. Hence, not only can a good market image and investor confidence be secured, but both social and economic benefits can also be expanded.

Ele.me sets an example with its fine-tuned interface

For implementation, enterprises are encouraged to identify low-cost ESG measures that yield high social benefits through experimentation. At the same time, companies can seek collaboration with academia to conduct precise assessments of the costs and social benefits of specific ESG action plans. We will outline a case of collaboration between Alibaba and academia to illustrate how businesses can derive greater social benefits at lower costs.

Ele.me, Alibaba Group’s online delivery services platform, is the second largest food delivery company in China, with over 700 million users in 2022. During a collaborative project with the platform, we studied how “green nudges” impacted the use of disposable tableware. Specifically, Ele.me has started a “green nudge” experiment in Beijing, Shanghai, and Tianjin. For customers in these three cities, the default option on the ordering interface is set to “no need for tableware” and those who choose this default option are awarded “Ant Forest ‘s Green Energy” points. This is a non-cash customer incentive. Once a customer has collected enough points, Alibaba will, in the name of the customer, plant trees in a desert area or launch other environmental protection actions.

Such a change may involve minimal costs for Ele.me but what social benefits can it bring? Our analysis of users’ orders in 10 major Mainland cities between 2019 and 2020 illustrates that cities where “green nudge” measures have been introduced have seen a 648% surge in no-cutlery orders (see Figure). Nationwide implementation of such measures is expected to save over 21.75 billion sets of single-use cutlery, thus reducing 3.26 million tons of plastic waste and saving 5.44 million trees from being cut down for timber. This study was featured as the cover story of the Science magazine in 2023, gaining wide attention from global media.

Figure    Share of no-cutlery orders in Ele.me’s green-nudge experiment: before and after

This case study demonstrates that it is possible for enterprises to honour their social and environmental commitments at a low cost. Just a few hours of work by a programmer is enough to generate tremendous social value. Such an innovative ESG action has not only boosted corporate ESG performance but also brought actual social benefits conducive to achieving national environmental goals.

Collaborative verification of strategy outcomes by enterprises and academia

While the collaborative study between businesses and academia mentioned above is just the tip of the iceberg, this methodology can be applied to the analysis of various problems. For instance, how can a leading company manage supply chains in terms of “E” in ESG? Given budget constraints, should enterprises invest more in reducing carbon emissions or focus more on air pollution management (in terms of “E” in ESG)? How would a wider diversity of staff and management impact the financial and ESG performance of businesses (in terms of “S” in ESG)? What assessment and evaluation mechanisms are most beneficial for enhancing business performance and staff satisfaction (in terms of “G” in ESG)? While it may be challenging for enterprises to find answers to these questions, it is a less daunting task for academia. By collaborating with academia, companies can leverage its theoretical base and data analytical capabilities to more precisely identify ESG opportunities and verify the effectiveness of their strategies. In our opinion, as far as ESG ratings are concerned, enterprises are not just “exam candidates” but should be drivers and practitioners of ESG. Undoubtedly, more collaborations between companies and academia will give a powerful impetus to ESG innovations.

Prof. Guojun He
Professor in Economics
Director, HKU Jockey Club Enterprise Sustainability Global Research Institute
Associate Director, Institute of China Economy

Ms Wendy Cui