MDDI 演講稿 · 2025-08-26

吳漢雁高階政務次長在 AI4Life 峰會 2025 的開幕致辭

· 演講者 · AI4Life 峰會

要點

  • 傳統藥物研發往往耗時超過10年、耗資數十億美元才能讓一種藥物抵達患者,AI端到端流水線正在大幅壓縮這一週期。
  • 南洋生物製藥的「藥靶相互作用圖神經網路」展示了AI如何縮短研發週期、降低成本並拓展未滿足醫療需求的治療可能性。
  • AI驅動的ADMET平臺與虛擬分子篩選技術,使研究人員能在早期預測藥物吸收、代謝與毒性,將高風險化合物在進入後續階段前提前過濾。
  • 新加坡以對A*STAR及學術醫學中心的持續投入為核心,藉助聯邦學習方法,允許研究人員在不傳輸敏感患者資料的前提下跨醫院協作開發AI模型。
  • 南洋生物製藥、NVIDIA、慧與科技(HPE)及Equinix在峰會上簽署諒解備忘錄,將在新加坡共建全球最大的AI驅動天然化合物庫。
  • 勞動力晉升聯盟正推行培訓計劃,為新加坡人提供資料科學、分子建模與臨床轉化等多學科技能,以應對AI生物醫藥領域的新興崗位需求。

完整譯文(繁體中文)

MDDI 英文原文譯文 · 翻譯日期: 2026-06-21

南洋生物製藥公司主席 Roland Ong 博士,

來自 NVIDIA 的各位貴賓,

Hewlett Packard Enterprise、Equinix 及勞動力發展聯合會的各位,

女士們、先生們,

下午好,

很榮幸今天出席 AI4Life 計劃的啟動儀式。這是一項意義重大的舉措,旨在藉助人工智慧為人類健康探索新的解決方案。

人工智慧創新,助力生物醫學科學發展

藥物研發歷來耗資巨大且充滿不確定性。一種藥物從研發到最終到達患者手中,往往需要超過10年時間和數十億美元的投入。對於急需幫助的人來說,這實在太漫長了。

然而,人工智慧正開始改變這一格局並取得初步成效。端到端的人工智慧驅動流程已產出藥物候選化合物,進入臨床試驗所需的時間僅為傳統方式的一小部分。南洋生物製藥公司的藥物-靶標相互作用圖神經網路(Drug-Target Interaction Graph Neural Network)等平臺展示了人工智慧如何縮短研發週期、降低成本,併為未被滿足的醫療需求拓展治療可能性。這項技術固然複雜,但真正重要的是它如何影響和改善我們的生活。因此,我非常感激你們正在踏上這段非凡的旅程。

人工智慧通過整合多層生物資料——從基因組學到蛋白質組學再到臨床洞察——為科學家提供了對疾病通路更深刻的理解。人工智慧還幫助科學家克服長期存在的瓶頸。虛擬分子篩選讓研究人員無需在實驗室篩查數百萬種化合物,而是專注於最有前景的候選物,從而節省了寶貴的時間、資金和其他資源。

除篩選之外,人工智慧也開始讓藥物開發變得更加安全。機器學習方法可以預測在研藥物在體內的行為,而人工智慧驅動的 ADMET 平臺則預測藥物的吸收與代謝方式及其是否具有毒性。這些平臺能夠在高風險化合物進入研發流程太深之前就將其篩除。這不僅提高了安全性,還降低了所需的成本和資源。

在人工智慧如何革新我們應對人類健康問題的方式上,我們才剛剛起步。仍有許多空白有待填補。例如,人工智慧設計的分子在合成上依然頗具挑戰,資料匱乏也是一大問題。

儘管如此,方向對我們所有人來說都是清晰的。從藥物研發到臨床試驗,人工智慧已經在發揮巨大作用。即便在日常醫療保健成效方面,人工智慧也將通過可穿戴裝置和移動健康應用程式等工具,實現更個性化的醫療、更早期的疾病檢測以及更好的患者參與。

強化新加坡作為全球醫療創新樞紐的地位

AI4Life 也體現了新加坡推動創新的方式——即通過與在場每一位的合作共同實現。這一計劃匯聚了眾多利益相關方:學術界、研究人員、全球科技企業,共同為大規模加速醫療創新提供一個範式。

我們相信新加坡具備充分條件來擔當這一引領角色。多年來,我們在生物醫學科學生態系統上進行了大量投資,其核心是對 A*STAR 及各學術醫療中心的持續投入。這些中心將患者護理、教育與研究融為一體,不僅提供高質量的醫療服務,也推動研究前沿向新加坡延伸。與此同時,我們可信賴的數字基礎設施為大規模人工智慧應用提供了堅實基礎,涵蓋醫療服務提供和支援轉化研究兩個層面。例如,利用聯邦學習方法,研究人員可以跨多家醫院協作開發人工智慧模型,而無需轉移敏感的患者資料。這既保障了隱私,又加速了疾病預測和個性化醫療領域的創新。

為此,我非常高興見證南洋生物製藥公司、NVIDIA、Hewlett Packard Enterprise 與 Equinix 之間諒解備忘錄(MoU)的簽署。這是一個四方聯盟,旨在新加坡建立全球規模最大的人工智慧驅動天然化合物庫。此次合作融合了深度技術、先進基礎設施與生命科學專業知識,以加速治療性藥物發現和生物醫學創新。

我們必須持續尋求合作途徑,大規模加速醫療創新。通過開展國際合作和跨行業協作,新加坡既可引入寶貴的專業知識、借力全球研究成果,也可通過技能發展和智慧財產權創造來獲取價值。這將確保我們的長期競爭力,並將新加坡定位為人工智慧驅動藥物研發的樞紐。

本月早些時候,政府啟動了經濟戰略檢討(Economic Strategy Review),以制定經濟藍圖。與 AI4Life 等計劃相結合,這些舉措確保人工智慧領域的突破轉化為切實影響,惠及患者、新加坡整體經濟和社會。我認為,這賦予了一切真正的意義——這絕不僅僅是為了技術而技術。

培育面向未來的生物醫學人才隊伍

最後,我想說的是,僅靠技術本身並不能帶我們到達目標。人工智慧與生物技術的融合正在資料科學、分子建模和臨床轉化領域催生新的職位。為抓住這些機遇,新加坡必須投資培育跨學科人才——能夠在人工智慧、生物學與藥物研發之間架起橋樑的人。這些技能組合對於將強大人工智慧演算法的預測轉化為現實療法至關重要。

我們已有相關計劃付諸實施,並將持續完善。勞動力發展聯合會的勞動力發展計劃正在為新加坡人賦予所需技能,以承擔這些新興職位。這確保我們的人才培育與行業發展保持同步,使我們擁有一支面向未來的人才隊伍,能夠支援創新,為新加坡人、新加坡乃至整個世界帶來更好的成果。

結語

總而言之,人工智慧在醫療健康領域的前景無限廣闊,我相信我們所有人都支援這一事業,這也正是我們今天聚集於此的原因。AI4Life 是我們將這一前景轉化為現實的有力範例。

現在,讓我們乘勢而上。讓我們攜手跨越學科、行業、企業和利益相關方群體的界限,共同合作,確保新加坡不僅緊跟時代步伐,更能引領醫療創新未來的走向。

非常感謝。

英文原文

MDDI 官網原始記錄 · 抓取日期: 2026-06-21

Dr. Roland Ong, Chairman of Nanyang Biologics,

Distinguished Guests from NVIDIA,

Hewlett Packard Enterprise, Equinix, and the Workforce Advancement Federation,

Ladies and Gentlemen,

A very good afternoon,

It is my pleasure to join you today at the launch of the AI4Life initiative. This is a significant initiative to harness AI to discover new solutions for human health.

AI Innovation for Better Biomedical Science

Drug discovery has traditionally been costly and uncertain. It often takes more than 10 years and billions of dollars before a single medicine reaches patients. That is really long when someone is in need of help.

However, AI is beginning to change this landscape and yield early success. End-to-end AI-driven pipelines have produced drug candidates that take a fraction of the time to reach clinical trials. Platforms like Nanyang Biologics’ Drug-Target Interaction Graph Neural Network show how AI can cut discovery timelines, reduce costs, and expand treatment possibilities for unmet medical needs. The technology is really complex but what really matters is how it affects and improves our lives. So I am very grateful that you are embarking on such an amazing journey.

AI offers scientists a deeper understanding of disease pathways by combining multiple layers of biological data — from genomics to proteomics to clinical insights. AI also helps scientists to overcome long-standing bottlenecks. Instead of screening millions of compounds in the lab, virtual molecular screening has enabled researchers to focus on the most promising ones, therefore saving precious time, money and other types of resources.

Beyond screening, AI is also starting to make drug development safe. Machine learning methods can predict how a drug in development behaves in the body, while AI-powered ADMET platforms – which predict how a drug is absorbed, processed, and whether it might be toxic. These platforms can filter out risky compounds early on, before they move too far down the pipeline. Again, it makes it safer and reduces the cost and resources required.

We are just at the starting block on how we see AI revolutionising the way we address human health. There are still many gaps to be filled. For example, AI-designed molecules can still be really hard to synthesise, and data scarcity is a problem.

That said, the trajectory is clear all of us. AI is already making a huge difference from drug discovery to clinical trials. Even in day-to-day healthcare outcomes, AI will enable more personalised medicine, earlier detection of disease, and better patient engagement through tools such as wearables and mobile health apps.

Strengthening Singapore’s Role as a Global Healthcare Innovation Hub

AI4Life also reflects how Singapore approaches innovation – and that is through partnerships with each and every one of you in the room. This initiative brings together many stakeholders: academics, researchers, global technology players, all here to offer a model for accelerating healthcare innovation at scale.

We believe that Singapore is well positioned to play this leading role. Over the years, we have invested heavily in our biomedical science ecosystem, which is anchored by sustained investments in A*STAR and our Academic Medical Centres. These centres integrate patient care, education, and research, enabling not just high-quality healthcare but also pushing the frontiers of research into Singapore. This is also then coupled with our trusted digital infrastructure, which provides a strong foundation for large-scale AI applications in both the provision of healthcare services and supporting translational research. For example, using federated learning approaches, researchers can then collaborate to develop AI models across multiple hospitals without having to transfer sensitive patient data. This ensures privacy while accelerating innovation in disease prediction and personalised medicine.

To this end, I am very pleased to witness the MoU signing between Nanyang Biologics, NVIDIA, Hewlett Packard Enterprise, and Equinix. This is a four-party alliance to build the world’s largest AI-powered natural compound library, right here in Singapore. The collaboration combines deep tech, advanced infrastructure, and life sciences expertise to accelerate therapeutic discovery and biomedical innovation.

It is crucial that we continue to find ways to collaborate and accelerate healthcare innovation at scale. By collaborating internationally and across industries, Singapore can bring in valuable expertise and leverage global discoveries, while also capturing value through skills development and creation of intellectual property. This will then ensure our long-term competitiveness and position Singapore as a hub for AI-enabled drug discovery.

Earlier this month, the Government launched the Economic Strategy Review to chart an economic blueprint. Together with initiatives like AI4Life, they ensure that breakthroughs in AI translate into real impact, for patients, for Singapore’s wider economy and for our society. I think that makes it all meaningful that it is not just technology for technology’s sake.

Building a Future-Ready Biomedical Workforce

And finally, I want to say that technology alone will not get us there. The convergence of AI and biotechnology is creating new roles in data science, molecular modelling, and clinical translation. To capture these opportunities, Singapore must invest in multidisciplinary talent - people who can bridge AI, biology, and drug development. These skillsets are essential in turning predictions from powerful AI algorithms into real-world therapies.

We have schemes in place to make this happen and will continue to improve. Workforce development programmes by the Workforce Advancement Federation are equipping Singaporeans with these skills to take on these emerging roles. This ensures that our talent development keeps pace with industry evolution, and that we have a future-ready workforce that can support innovation that can lead to better outcomes for Singaporeans, Singapore and the world at large.

Conclusion

In conclusion, the promise of AI in healthcare is immense, and I believe all of us are behind this cause, which is why we are here today. AI4Life is a powerful example of how we translate this promise into reality.

So now, let us build on this momentum. Let us work together across disciplines, sectors, companies, stakeholder groups, to ensure that Singapore not only keeps pace, but leads in shaping the future of healthcare innovation.

Thank you so much.