MDDI 演讲稿 · 2025-11-26

政务部长刘燕玲在AIMX新加坡的开幕主旨演讲

政务部长刘燕玲在AIMX新加坡的开幕主旨演讲

Jasmin Lau · MDDI 政务次长 · AIMX 新加坡

要点

  • 新加坡科技与创新经济战略审查委员会由政务部长 Jasmin Lau 与政务次长 Goh Hanyan 联合领导,正围绕「应用」「整合」「信任」三大支柱制定国家 AI 战略框架。
  • 在「应用」层面,新加坡的战略重心是在各行各业全面落地 AI,而非自建大型语言模型,并要求企业从零出发构想「AI 原生」的行业形态,而非仅追求边际效率提升。
  • 在「整合」层面,委员会主张建设开放基础设施,让全球各规模企业均可获取算力、数据与 AI 模型,确保 AI 机遇不因财力门槛而受限。
  • 「整合」支柱还呼吁设立更多测试床与沙盒,推动有前途的 AI 试点走出「沙盒炼狱」、升级为全国性解决方案,并通过开放协作将全球 AI 模型本地化以服务新加坡与区域市场。
  • 在「信任」层面,新加坡计划将本国标准嵌入数字经济协议及国际合作框架,使在新加坡诞生的产品与服务在全球范围内获得公信力认可。
  • 「信任」支柱还包括将 AI Verify 和 AI Assurance 工具作为「信任即服务」向全球企业开放,供其对照新加坡标准验证 AI 系统,并设立「新加坡信任标志」以标示可靠性、安全性与品质。
  • 委员会指出,心态转变是新加坡面临的最大挑战,需在全国层面从规避风险的「先行者劣势」思维,转向更具好奇心、勇气与敏捷性的 AI 采用方式。

完整译文(中文)

MDDI 英文原文译文 · 翻译日期: 2026-06-21

大家早上好。

感谢今天邀请我出席第三届 AIMX Singapore。

今天在座的许多人,大概是想了解 AI 如何能帮助你们的行业或公司。

或者你们是想了解其他人如何运用 AI 助力公司发展,同时摸清竞争态势。

你们中的大多数,今年至少参加过一场其他 AI 或科技会议,也许不止一场。有时候,我觉得唯一比 AI 增长更快的,就是 AI 会议的数量。

就在过去这一年里,我参加了许多这样的科技会议——但我一点也不觉得厌倦。每个月去参加新的会议,都能看到新的进展——新工具、新技术。

技术发展非常迅猛,但当你听到如此之多的观点时,可能会深感应接不暇。

那该怎么办呢?你可以转向 AI,尝试问它:"你能为我推荐一些 AI 时代的商业策略吗?"或者"你能建议我哪家 AI 解决方案供应商最适合我的公司吗?"

现在,请设想一下,如果 AI 在新加坡 1965 年刚刚独立时便已存在,而我们的开国领袖在那时向 AI 询问如何建设经济——它会给出什么答案?

我在 ChatGPT 中输入了以下提示词:"鉴于新加坡 1965 年的经济状况和政治环境,根据经济学文献与研究,当时应推荐怎样的经济战略?"

AI 给出的回答是:

它会建议新加坡明哲保身,保持谨慎。

用关税和配额保护本国产业。

保持一个低调的贸易港口定位,专注于低成本制造业。

依赖更大的邻国获取市场规模。

避免在跨国企业上押下豪赌。

但 AI 无法预见、也看不到的,是我们这个小国敢于迈出不可思议之跨越的能力。

70 年代,我们将裕廊从沼泽之地建成工业城镇——从一个简单的贸易站蜕变为举足轻重的制造业中心。

80 至 90 年代,我们决定与全球市场接轨,将自身打造为区域金融中心和航运枢纽。

进入 2000 年代,我们投资研发,大力发展生物医药产业,并立志成为全球科技中心。

这一切,都不是 AI 会推荐的方向。

今天,我们正在迈入我们现在所称的"智能经济"时代。

我们看到 AI 能够如此大幅地放大人类能力,使得小团队能以前所未有的速度打造独角兽企业。

但这也给我们带来了令人不安的局面。是的,AI 拉平了整个经济领域的竞争起跑线。这同时意味着,曾经让新加坡脱颖而出的优势,如今已更容易被其他国家和城市复制。

AI 让学习我们最佳实践变得轻而易举。

随着越来越多的国家认识到 AI 的潜力,这也意味着许多国家和企业将重构供应链,寻求更安全的据点。

因此,新加坡政府当前面临的问题是:在这个新世界中,我们如何保持自身的相关性、竞争力与信心?

信心至关重要,因为我们的人民需要对政府引领新加坡走过这一新时代的方式充满信心。

我们设有一个负责科技与创新领域的经济战略检讨委员会,由我与高级政务次长吴汉英共同领导。我的部分委员会成员今天也在场——丽怡和天怡。感谢二位的出席。

我们目前还没有所有答案,但我认为这是一个分享我们初步思考的好机会。我们今天也希望听听你们的想法——如果你们有任何反馈,随时给我们留言。

我们围绕三个领域来整理思路——应用、整合与信任。

首先是应用。新加坡长期以来以把事情做好著称——有条不紊、可靠高效。但AI不只是奖励效率,它奖励那些愿意尝试的人,这意味着要有勇气去做新的事情。AI奖励那些愿意想象此前不可能之事的人。

因此,我们的领先行业和企业现在必须提出更令人不安的问题。他们不应该问:「AI能让我们提速10%或20%吗?」而应该问:「如果我们今天从零开始,在一个AI原生的世界里,这个行业会是什么样子?」

当我们谈论应用时,这不是要构建下一个大型语言模型(Large Language Model),因为我们认为这不是新加坡最能发挥优势的地方。我们认为,重点在于将AI应用于新加坡的每一个行业和领域。

我想举几个例子。在物流领域,我们是否只是在尝试实现港口运营的自动化?或者,新加坡能否成为全球的智能指挥中心——供应链能够因中断、天气或需求激增而即时调整?系统能够在任何人察觉之前自行检测故障并修复?

听起来有点让人担忧,也难以实现,但这正是我们希望开始聚焦的愿景。

在医疗健康领域,我们是否仅将AI用于改善诊断?或者我们能否构建预测性和预防性的医疗模型,不仅延长寿命,更延长健康寿命?

这些都是值得提出的问题。我今天对在座各位的期望是:当你们参观展览时,问问自己——在你们自己的公司里,在这个AI世界中,你们会做哪些完全不同的事情?

这些转变不仅需要新的技术工具,还需要新的思维方式和新的合作关系。但正是这些雄心壮志,将使新加坡保持领先——不是把同样的事情做得更好,而是去做全新的事情。

第二是整合。新加坡长期以来的优势在于让系统的不同部分协同运作。政府、企业、学术界、研究机构——基于国际标准,我们非常善于协调。

但在当今的智能经济中,我们需要将新加坡视为一个开放的网络,而不仅仅是单一的生态系统——一个来自世界各地的最佳创意可以落地、生长和规模化的地方。

我们的委员会正在思考如何通过三种方式尽可能保持新加坡的开放性。

开放基础设施。我们的算力、数据和AI模型必须让各种规模的企业都能获取——无论来自世界何处。AI的机遇不应仅限于那些负担得起的人。

开放实验。我们在政府、金融和医疗健康领域已经推出了许多测试平台和沙盒。但我们还需要更多。

在这些沙盒中,企业可以安全地试验AI解决方案,有潜力的创意不会困于"试点炼狱"。有些AI项目在试点阶段停滞太久,都该获得长期服务奖了。

我们需要认真思考如何将试点推进为国家级解决方案、如何有意义地扩展规模,以及如何确保沙盒仍然存在,但要留给新的试点项目。

开放协作。我们或许不会构建世界上最大的AI模型,但我们可以汇聚研究人员、初创企业、导师和政府,将来自世界各地的这些模型加以适配,使其在新加坡以及整个地区运行得更好、更快、更安全。

委员会重点关注的第三个领域是信任。

在AI可以被部署于任何地方的时代,信任是那项难以被轻易取代的稀缺资产。

建立信任花了我们数十年。对我们机构的信任、对我们治理的信任,以及对"新加坡说某件事会奏效,它就真的会奏效"的信任。

但信任也必须与时俱进。当我们走得太慢,信任就变成了自满;当我们走得太快,信任就会变得非常脆弱。我们都知道,一旦失去信任,要重建是极为困难的。

我们的委员会相信,我们能够将信任转化为新加坡的战略优势。我们希望信任成为我们的贸易货币。我们可以通过数字经济协议(Digital Economy Agreements)和国际合作,将新加坡的标准嵌入全球协议。

未来,来自新加坡的产品和服务将在全球各地获得信赖。

我们还希望提供"信任即服务"。AI Verify或AI Assurance等工具将使全球企业能够依据透明、严格、由新加坡制定的标准来验证其系统。

我们希望将信任提升为我们的溢价标签。在这里构建的产品和服务应当带有"新加坡信任标志"(Singapore trust mark)——这是可靠性、安全性和质量的信号。

但我们更大的挑战,是思维方式。

新加坡人善于谨慎和自律,但AI奖励那些充满好奇、勇于探索、灵活敏捷的人。

我们确实需要适应那片灰色地带,适应那种模糊性。我们确实需要承担一些经过计算的风险,并且必须接受并非一切事情都能一次成功。

上周有几位企业领导人在一次对话中告诉我:「你知道面对先行者劣势是什么感觉吗?我可以率先购买这项技术并尝试推行,但如果我第一个出手,损失可能很大。我想让别人先试,如果成功了,我再跟进。如果失败了,还好我没做。」

这正是我们需要克服的那种思维方式。我们并不是说新加坡每个人都必须成为先行者,但作为一个国家,我们确实需要更多这样的思维方式。

我们还需要这样一类人才——他们不只懂得如何使用 AI,更能够想象出解决问题的全新方式。

我们曾多次自我革新——从贸易站到制造业中心,再到全球枢纽。如今,我们面临的挑战是迈出下一个跨越:从高效走向创新,从善于治理走向塑造世界,从遵循最佳实践走向制定最佳实践。

如果我们能做好这一点,新加坡将不只是智能经济的参与者,更将以我们的理念与想象力,共同塑造这一经济形态。

最后,祝本次活动圆满成功。请大家沉浸于今日展示的各项技术之中,并挑战自我,力争成为先行者,退而求其次也要做第二个行动者。

谢谢。

英文原文

MDDI 官网原始记录 · 抓取日期: 2026-06-21

Good morning, everyone.

Thank you for having me here today at the 3rd edition of AIMX Singapore.

Many of you are here today, probably because you want to find out how AI can help your industry or company.

Or maybe you are keen to find out how others are using AI to help their companies, and you are trying to suss out the competition.

Most of you would have attended at least one other AI or technology conference this year, perhaps more. Some days, I feel the only thing growing faster than AI is the number of AI conferences.

Just in the past year alone, I’ve been at many of these technology conferences – but they don’t make me feel bored. Every month, when you go for a new conference, you see new developments being displayed – new tools and technologies.

The technology is moving very fast, but it can get very overwhelming when you hear so many perspectives.

So what do you do? You could turn to AI and try to ask, “can you recommend me some business strategies in the age of AI?” or “can you suggest to me which AI solution vendor is most suitable for my company?”

Now imagine, for a moment, if AI had existed when Singapore first became independent in 1965, and our founding leaders had asked AI at that point, how to build our economy – what would it have said?

I put the following prompt into ChatGPT: “Given Singapore's economic situation and political environment in 1965, what would have been the recommended economic strategy for Singapore based on economic literature and research?”

The response from AI was that:

It would have urged Singapore to play it safe. Stay cautious.

Protect your industries with tariffs and quotas.

Remain a modest trading port. Focus on low-cost manufacturing.

Depend on your larger neighbours for market scale.

Avoid bold bets on multinational firms.

But what AI would not have predicted, or seen, was our capacity as a small country to take improbable leaps.

In the 70s, we turned Jurong from swampland into an industrial town – we shifted from a simple trading post into a serious manufacturing centre.

In the 80s and 90s, we decided to connect with global markets and built ourselves up as a regional financial centre and shipping hub.

And in the 2000s, we invested in research and development, built up our biomedical industry and we aimed to be a global tech hub.

All of this was not what AI would have recommended.

Today, we are entering what we now call the intelligence economy.

We see how AI can amplify human ability so dramatically that small teams can build unicorns faster than ever before.

But this also creates an uncomfortable situation for us. Yes, AI has levelled the playing field across the economy. It also means that what once made Singapore exceptional is now easier for other countries and cities to replicate.

AI makes it simple to learn our best practices

As more countries recognise AI’s potential, it also means that many of these countries and companies will restructure supply chains and seek safer bases.

So the question that our Singapore Government is now posed with is: How do we stay relevant, competitive and confident in the new world?

Confidence is very important, because our people need to feel confident in the way that the Government leads Singapore in this new era.

We have an Economic Strategy Review committee on Technology and Innovation, which I lead together with Senior Parliamentary Secretary, Goh Hanyan. Some of my committee members are here with me today – Lai Yee and Tianyi. Thank you for joining us.

We don’t have all the answers yet, but I thought it would be a good chance to share our early thinking. We also want to hear your ideas today – if you do have any feedback, drop us a note anytime.

We are organising our thoughts around three areas – Application, Integration, and Trust.

First, application. Singapore has long been known for doing things well – systematic, reliable, efficient. But AI doesn’t just reward efficiency. It rewards people who are willing to experiment, and that means having the courage to do something new. AI rewards people who are willing to imagine what was previously impossible.

So, our leading sectors and companies must now ask more uncomfortable questions. They shouldn’t be asking: “How can AI make us 10% or 20% faster?”. But they need to be asking: “If we were starting from scratch today, what would this industry look like in an AI-native world?”

When we are talking about application, it’s not about building the next Large Language Model, because that's not where we think Singapore could best play. We think it's about applying AI in every industry and sector in Singapore.

I thought it would be useful to give you some examples. In logistics, are we simply trying to automate our port operations? Or could Singapore be the world’s intelligent command centre – where supply chains can be adapted instantly due to disruptions, weather, or demand spikes? Where systems detect faults and fix themselves before anyone notices?

Sounds a little bit scary and impossible, but that is the kind of vision that we want to start focusing on.

In healthcare, are we using AI only to improve diagnostics? Or are we building predictive and preventative care models that extend not just lifespans but health spans?

A lot of questions worth putting out there, and my hope for all of you today is as you walk through the exhibitions, ask yourselves, n your own companies, what would you be doing completely differently in this AI world?

These shifts require not just new technology tools but new mindsets and new partnerships. But these are the ambitions that will keep Singapore ahead — not by doing the same things better, but by doing completely new things entirely.

Second, integration. Singapore’s long-standing strength is getting different parts of our system to work together. Government, business, academia, research – we are quite good at coordination, based on international standards.

But in today’s intelligence economy, we need to think of Singapore not as a single ecosystem, but as an open network. A place where the best ideas from around the world can land, grow, and scale.

Our committee is thinking about how to keep Singapore as open as possible, in three ways.

Open infrastructure. Our compute, data, and AI models must be accessible to companies of all sizes – from all around the world. AI opportunity shouldn’t be limited to those who can afford it.

Open experimentation. We have launched many testbeds and sandboxes in government, finance, and healthcare. But we need a lot more of these.

Sandboxes where companies can trial AI solutions safely, and where promising ideas don’t get stuck in pilot purgatory. Some AI projects get stuck in pilots for so long, they should qualify for long service awards.

We need to think harder about how to move pilots into national solutions, how to scale them meaningfully, and how to ensure that the sandbox still exists, but for new pilots.

Open collaboration. We may not build the world’s biggest AI models. But we can bring together researchers, startups, mentors, and government to adapt these models from around the world, and make them work better, faster, and safer for us in Singapore, and in the region.

The third area that the committee has been focusing on is trust.

In an age where AI can be deployed anywhere, trust is the rare asset that cannot be replaced as easily.

It has taken us decades to build. Trust in our institutions, trust in our governance, and trust that when Singapore says something will work, it does.

But trust too, must evolve. When we move too slowly, trust becomes complacency. When you move too fast, trust becomes very fragile. We all know that when we lose trust, it’s very hard to recover.

Our committee believes we can convert trust into our strategic advantage for Singapore. We want trust to be our trade currency. We can embed Singapore’s standards into global agreements through Digital Economy Agreements and international collaborations.

In the future, products and services coming out of Singapore will be trusted all over the world.

We also want to offer trust as a Service. Tools like AI Verify or AI Assurance will allow companies worldwide to validate their systems against transparent, rigorous, Singapore-developed standards.

We want to elevate trust as our premium. Products and services built here should carry the “Singapore trust mark” – a signal of reliability, safety, and quality.

But our bigger challenge, is mindset.

Singaporeans are good at being careful and disciplined, but AI rewards those who are curious, brave, and agile.

We do need to get comfortable with that grey space, that ambiguity. We do need to take some calculated risks, and we must accept that not everything will work on the first try.

Some company leaders told me last week during a dialogue: “Do you know how it feels to face first mover disadvantage? I could be the first one to buy the technology and try and implement it, but I lose big if I go first. I want someone else to try it, and then if it works, I'll do it. And if it doesn't, thank goodness, I didn't do it”.

That's the kind of mindset that we do need to get over. We're not saying everybody in Singapore must be a first mover, but as a country, we do need a bit more of that mindset.

We also need people who don't just know how to use AI, but who can imagine new ways to solve problems.

We've reinvented ourselves many times before – from trading post to manufacturing centre to global hub. Now, our challenge is to make the next leap forward. From being efficient to inventive. From being well-run to world-shaping. From following best practices to setting them.

If we can do this well, Singapore will not just be a participant in the intelligence economy. We will help shape it with our ideas and imagination.

On that note, I wish you a fruitful event. Please immerse yourselves in the technologies on display today and challenge yourselves to be the first mover, if not the second.

Thank you.