appConfigEntityId

bodyType

customContents

entityId

entityUuid

isVideo

articlePageQuery

article

articleSchemaContent

content

hooksArticlePageAdvertisementConfigArticle

hooksArticlePageFloatingButtonConfigArticle

appConfig

filter

json

liveArticle

Letters

Feel strongly about these letters, or any other aspects of the news? Share your views by emailing us your Letter to the Editor at letters@scmp.com or filling in this Google form. Submissions should not exceed 400 words, and must include your full name and address, plus a phone number for verification.
While generative artificial intelligence models such as OpenAI’s ChatGPT have made immense progress in recent years, some people – especially non-technical users – still question the efficiency of such technology.
In fact, the actual breakthrough of ChatGPT is its user-friendly interface, which allows even people without programming skills to have conversational interactions with the AI model.
For its language models to be of more practical use than as a chat pal, OpenAI allows developers to fine-tune GPT-3, a version of ChatGPT, according to the attributes of particular applications, resulting in more accuracy and more pertinent answers.
Generative AI can be incorporated in multiple sectors, including e-government, healthcare, finance, education and entertainment. It has a role in augmenting production efficiency, heightening user experience and creating additional economic value. Consider the domain of education, as an example. A fine-tuned model will be able to generate individualised educational material, tasks and examination plans depending on students’ learning style, abilities and interests, thus amplifying their participation, understanding and ability to retain information.
Additionally, such a platform can be adjusted to suit different learning trajectories, with specific levels of difficulty, thereby optimising learning results and stimulating curiosity.
To fully realise the potential of AI models like ChatGPT, developers must fine-tune them to devise ever more specialised and productive applications that meet user needs.
Furthermore, China should step up efforts in the research and development of generative AI technology to strengthen the operational efficiency of businesses and to foster social advancement. Besides, the development of China’s own generative AI models that are as good as ChatGPT is essential to the nation’s future success.
However, developing such technology requires ample computing resources and data sets, as well as high-performance computers and graphics processing units. For this reason, investing in the necessary chip technology is important, and this is an area where Hong Kong can help.
All in all, generative AI technology is of enormous significance to the development of mankind. China could strive to take the lead in the area of generative AI, thereby helping to elevate the country’s economic well-being and to improve people’s lives.
Leonard Chan Tik-yuen, vice-convenor, professional affairs committee (technology innovation), Liberal Party
Once upon a time, America loved a downtrodden China
Three cheers to your correspondent for “Containing China is a waste of West’s energy” (March 21). I cannot agree more. China’s rise is unstoppable, regardless of whatever people might wish. The rise is fast, and happening before our eyes.
The Chinese at large do not hate the Americans, though they are baffled by the turn of events. America looked so benign when it helped educate dozens of young Chinese boys in the 1870s. Among them was Zhan Tianyou, one of the pioneers of Chinese railroading, who trained at Yale.
Also, at the turn of the last century, the United States used the Boxer Indemnity (also known as the Gengzi Indemnity, according to the Chinese reckoning of the years) to build the predecessor to Tsinghua University, now one of China’s top universities. Must China be utterly downtrodden and defenceless before the US will love the Chinese?
K.M. Lau, Mid-Levels
Short sellers not to blame for the SVB crisis
I read with bemusement and frustration a contribution to the Opinion section, “In wake of SVB, Credit Suisse and other bank troubles, should hedge funds be banned to avert financial meltdown?” (March 21), where the author seems to lay the blame for the recent banking crisis at the feet of hedge funds in contradiction to well-established facts on the matter.
As readers will likely know, Silicon Valley Bank’s demise was brought about by the acute sensitivities that both its assets and liabilities had to rising interest rates, that is to say a run on the bank, combined with the sharp decrease in value of its portfolio of long-dated, fixed-term bonds. Short sellers may well have held positions on the bank, but to blame them for its collapse is, respectfully, absurd. Indeed, short sellers were among the first to make the public and regulators aware of their concerns about SVB and Signature Bank.
The article rehashes tropes about short sellers which have been debunked by regulators, academics and other stakeholders. In particular, the author references and calls for a revisit to the short-selling bans that came in the aftermath of the 2008 financial crisis. However, at least one of the regulators that enacted these bans – including Britain’s Financial Conduct Authority, the US Securities and Exchange Commission and the European Securities and Markets Authority – subsequently admitted that the bans did not have their intended effect. Moreover, it has been noted that short-selling bans exacerbate periods of volatility by suppressing market liquidity and price discovery. Tellingly, studies point to markets underperforming where the bans are in place.
In closing, I would note that short sellers, along with the mechanics of securities lending that facilitates short shelling, promote liquidity and price transparency as well as accountability across capital markets by publishing in-depth research and exposing fraud.
Michael Bugel, co-head of Asia-Pacific, The Alternative Investment Management Association


Senior HSBC bankers face double blow as school-fee subsidy cut adds to job risks: sources

Hong Kong talent schemes could add 1.2% to GDP annually: John Lee

Hong Kong IT analyst charged with murdering wife at their flat

China launches Pinglu Canal; Hong Kong unveils first 5-year plan: SCMP’s 7 highlights

Hong Kong

ChatGPT can be fine-tuned to serve us all better

A person chats with ChatGPT in Chinese in Shanghai on February 6. Photo: Getty

How universities can start to grapple with ChatGPT’s capabilities

The rise of the AI-powered chatbot ChatGPT has caught the education sector off guard. Image: Shutterstock 

How academia can embrace ChatGPT and reignite a love for learning

ChatGPT could be seen as a tool, like the calculator or Microsoft Excel, taking up automated low-level, repetitive tasks and allowing us to get into higher-level, more complex thinking, and to be more creative. Photo: Shutterstock

Opinion

Readers discuss how to fully realise the potential of ChatGPT, happier times in US-China relations, and short sellers’ role in capital markets


Readers discuss how to fully realise the potential of ChatGPT, happier times in US-China relations, and short sellers’ role in capital markets.


ChatGPT and other generative AIs

Artificial intelligence (AI)

US-China relations

Silicon Valley Bank (SVB)

.css-1mniedq{color:inherit;font-weight:inherit;font-size:inherit;font-family:inherit;line-height:inherit;}Readers discuss how to fully realise the potential of ChatGPT, happier times in US-China relations, and short sellers’ role in capital markets

Readers discuss how to fully realise the potential of ChatGPT, happier times in US-China relations, and short sellers’ role in capital markets