Members of Singapore's literary community are calling on the National Library Board of Singapore, or NLB, to exercise greater prudence in adopting generative artificial intelligence or risk "permanently damaging Singapore's literary landscape".

A collective statement signed by 68 writers released on Jan 7 questioned the NLB's "uncritical endorsement" of the technology. The library has introduced a series of programs using AI since January 2024, including generative AI prototypes for immersive experiences.

The signatories — who include writers, publishers, educators and other cultural workers — called on the library to remove any suggestion that generative AI is "an adequate substitute for traditional writing development". It also called on the NLB to educate the public on the technology's limitations, as well as its negative impact on learning and the environment.

This is the first collective statement by Singapore's literary community on the impact of generative AI on the writing landscape.

In April 2024, individual writers and publishers objected to the Infocomm Media Development Authority's plans to build a Southeast Asia-focused large language model, but stopped short of a collective stance.

This latest statement also cited generative AI's threat to a writer's intellectual property as one of the literary community's major concerns, adding: "NLB's promotion of AI has not been accompanied by warnings about the ethical problems of the field, and thereby normalizes intellectual theft."

Citing an NLB event titled Children Write: Publish A Book With Gen-AI, designed for participants aged 7 to 12, the statement raised concerns that such an event "furthers a belief that use of this technology can be a substitute for traditional writing skills".

It added that AI will adversely affect the quality of literature produced and that the technology's environmental costs — composing a single email with ChatGPT has been found to consume over half a liter of water — contradicts the library's sustainability initiatives.

The statement, addressed to NLB Chairman Lee Seow Hiang, Chief Executive Ng Cher Pong and Chief Librarian Gene Tan, as well as Minister for Digital Development and Information Josephine Teo, called for a consultation with members of the literary community to address these issues.

It acknowledged the relevance of AI tools and their potential applications in the literary arts, but added: "As a national institution, NLB is uniquely positioned to educate the public on how it is possible to use AI responsibly."

Making efforts

Author Ng Yi-Sheng, one of the organizers of the statement, wrote on publishing platform Substack: "Chief librarian Gene Tan has already read the letter and initiated dialogue."

An official NLB website has been used to educate users about the ethical issues of AI. "I personally find this insufficient," Ng added.

Concerns have escalated in recent years as models such as ChatGPT have been trained on materials without authorization from copyright holders.

In 2023, the names of Singaporean writers such as novelists Balli Kaur Jaswal, Ovidia Yu and Rachel Heng were found in the Books3 dataset, which lists thousands of authors whose copyrighted works have been stolen to train large language models similar to ChatGPT.

The NLB said the library values the trust of the literary community and its AI prototypes are meant to complement, not replace, authors' efforts.

The NLB said it would work more closely with the literary community to address ongoing concerns and create new opportunities for writers and publishers.

THE STRAITS TIMES, SINGAPORE