TL;DR – What You Need to Know in 30 Seconds

• AI isn’t one monolithic market—it’s three interconnected segments:

• 1. Pre-GenAI (traditional AI): Fundamental techniques that underpin data-driven solutions.
• 2. AI Training Market: Resource-intensive frontier models driving the next AI breakthroughs.
• 3. Enterprise AI Market: Real-world applications delivering measurable business outcomes.

• Understanding their interplay is critical for Asian businesses aiming to maximise ROI from AI investments.

Are We Missing the Bigger Picture in the AI Race?

From smarter chatbots to insightful analytics, AI’s not one market—it’s three interconnected ones, each shaping how Asia leverages technology.

If you’ve spent any time recently skimming headlines about artificial intelligence, you’d be forgiven for thinking that generative AI is the only show in town. But AI isn’t just ChatGPT, Midjourney, or flashy avatars of celebrities endorsing your new favourite tech gadget. Behind the scenes, three distinct but intertwined markets are at play: the Pre-GenAI Market, the Training Market, and the Enterprise AI Market.

But what exactly are these three markets, and why should Asian businesses care?

Let’s unpack them one by one and understand how they converge to drive the future of innovation across Asia.

1. The Pre-GenAI Market: The Building Blocks of AI

Generative AI may be the current media darling, but the roots of AI go far deeper. We’re talking about traditional AI—technologies like machine learning (ML), reinforcement learning, and computer vision. These foundational techniques have been quietly evolving for decades, long before ChatGPT ever typed out its first response.

Contrary to popular belief, traditional AI hasn’t lost its relevance—far from it. In fact, the rise of generative AI has amplified its importance. Why? Because generative AI feeds on data often produced by traditional AI methods. For instance, Dell Technologies frequently uses machine learning to streamline supply chains or improve factory efficiency. These methods don’t get less important just because GPT-5 is around the corner—they become essential.

In short, traditional AI is like rice in Asian cuisine—fundamental, reliable, and always necessary, no matter what fancy new dish appears on the menu.

2. The Training Market: Powering AI’s Frontier

Next up is the AI training market—think of it as AI’s heavy lifting division. This market is dominated by big names you’ll recognise (OpenAI, Google DeepMind, Nvidia, Meta) who are making gigantic investments in infrastructure to create foundational AI models. Picture rows and rows of servers, massive GPU clusters, and sprawling data centres, humming 24/7.

These frontier models—like GPT-4 or Gemini—require immense computational resources. This isn’t just about bragging rights; it’s about pushing the boundaries of what AI can do. The innovations here spill directly into practical tools businesses use every day, like AI-driven coding assistants or creative platforms for content creation.

In Asia, we’re seeing heavy investment in this market too. Take Singapore’s AI supercomputing initiatives or China’s Baidu and Alibaba building mega-AI clusters. These moves aren’t just technological vanity—they’re strategic investments in the future.

3. The Enterprise AI Market: Real-World Results

And then there’s the enterprise AI market, arguably the most pragmatic of the three. Enterprises aren’t racing to build the next ChatGPT killer. Instead, they’re laser-focused on AI that solves real business problems—like optimising inventory management, enhancing customer support, or boosting marketing effectiveness.

Unlike the flashy training market, the enterprise market moves slower but deliberately. Enterprises demand reliability, compliance, and measurable outcomes—exactly the opposite of the ‘move fast and break things’ mentality we see in frontier AI research.

Across Asia, the enterprise AI market is thriving precisely because it offers clear returns. Banks in Indonesia deploy AI-driven chatbots to handle customer queries efficiently. E-commerce giants in Vietnam and Thailand integrate predictive analytics to forecast inventory and customer demand. It’s AI that’s practical, measurable, and directly linked to ROI.

How These AI Markets Interconnect

Here’s the real takeaway: These three markets aren’t isolated islands; they’re deeply interconnected ecosystems.

Traditional AI gathers and prepares the essential data. The training market produces foundational AI models and cutting-edge tech innovations. Enterprises then integrate both, using these tools and data to transform operations and customer experiences.

Think about it this way: traditional AI builds the roads, the training market crafts powerful engines, and the enterprise market drives the cars, delivering real-world value. Without any one of these, the system falters.

For instance, enterprises use AI-powered data agents to analyse massive datasets prepared by traditional AI methods. They then leverage frontier AI models (like generative AI) trained in data centres to extract actionable insights. The whole system is interdependent—each component driving progress in the other.

Why Does This Matter to Asia?

Asia is a unique melting pot of digital maturity, economic growth, and competitive intensity. Understanding these three markets isn’t just academic—it’s crucial for businesses looking to harness AI’s full potential.

For instance, enterprises in Southeast Asia’s rapidly expanding digital economy (expected to hit $263 billion GMV by 2025 according to Google’s recent e-Conomy SEA 2024 report) need practical AI solutions that deliver immediate business value. On the other hand, countries like Singapore, South Korea, and Japan are leading investments into the training market, building the infrastructure needed to power Asia’s next generation of AI innovations.

Simply put, knowing how these three AI markets interact helps Asian businesses invest smarter, act faster, and innovate effectively.

As we look ahead, Asia is uniquely positioned to benefit from understanding this AI ecosystem deeply. Whether you’re in manufacturing, finance, e-commerce, or healthcare, your business will inevitably interact with all three markets—whether you realise it or not.

Now, here’s something for you to ponder (and comment below!):

Which of these AI markets do you think will dominate Asia’s tech landscape by 2030? Will traditional methods endure, frontier models take over, or will enterprise solutions reign supreme?

We’d love to hear your thoughts.

You may also like:

• A Cautionary Tale for Asian Enterprises
• Accenture and Nvidia’s AI Power Play in Asia
• Perplexity’s Deep Research Tool is Reshaping Market Dynamics
• Learn more by tapping here.

TL;DR – What You Need to Know in 30 Seconds

• Grab CEO Anthony Tan believes workers and companies that don’t embrace AI risk being replaced by those who do.
• Grab paused normal operations for a nine-week generative AI sprint, significantly boosting innovation.
• AI tools developed by Grab, such as driver and merchant assistants, are empowering everyday entrepreneurs.
• Globally, many companies are downsizing due to AI, but Tan insists AI enhances human capabilities rather than replacing them.

Is Your Refusal to Embrace AI Secretly Sealing Your Fate?

Anthony Tan, co-founder and CEO of Grab—the Southeast Asian super-app that transformed regional transport, food delivery, and financial services—has made a bold and slightly unsettling prediction: “Humans who don’t embrace AI will be replaced by humans who embrace AI.”

In other words, whether you’re a company or an individual, ignoring AI isn’t merely shortsighted—it’s career suicide.

But before we panic, what exactly does Tan mean?

Making Humans ‘Superhuman’

Speaking at Converge Live in Singapore, Tan explained to CNBC’s Christine Tan that AI isn’t just a fancy tech upgrade. Instead, it’s a crucial tool to “make you superhuman” by significantly boosting productivity and freeing up valuable time.

Tan himself isn’t just preaching—he’s practising. Despite not being a coder, he’s enthusiastically using AI coding assistants for personal and professional projects. He claims AI has radically changed his productivity, helping him accomplish things previously impossible.

I can’t code myself, but I use AI to build my own projects, for research, for Grab,” Tan explained. “It totally changes how you spend your time.

Grab’s Radical AI Experiment

Grab didn’t stop at encouraging individual AI adoption. Instead, the company took it to a whole new level, implementing an ambitious, company-wide nine-week “generative AI sprint”.

This meant putting all regular business on pause to explore AI-driven solutions across the entire company. As Tan humorously admitted:

People thought I was crazy—maybe I am—but it really moved the needle.

During this sprint, Grab developed powerful AI tools, including:

• Driver Co-pilot: An AI assistant reducing wait times and boosting job opportunities for drivers.
• Merchant AI Assistant: Imagine a single mother in Jakarta now equipped with an AI-driven sous chef, packaging expert, and even a chief revenue officer—all in one assistant. This innovation isn’t just about efficiency; it’s empowerment, reshaping the livelihoods of Grab’s vast network of entrepreneurs.

The Wider Implications for Asia

This isn’t just a Grab-specific phenomenon. According to the World Economic Forum’s 2025 Future of Jobs Report, 40% of employers globally plan to downsize due to AI, and a whopping 86% anticipate AI reshaping their businesses by 2030.

Asia, in particular, with its digitally fluent workforce and vibrant entrepreneurial scene, stands uniquely poised to lead this transition. Grab’s aggressive AI strategy under Tan’s leadership could become a model for businesses across Southeast Asia, showcasing how AI can be harnessed responsibly and productively.

Human vs AI: Not a Zero-Sum Game

Tan stresses AI shouldn’t evoke fear—it should inspire excitement. AI adoption isn’t about machines replacing humans. It’s about humans becoming irreplaceable by effectively harnessing these tools.

If you’re reluctant or sceptical, Anthony Tan’s message is clear: embrace AI now, or watch as those who do leave you behind.

Hot Take

Anthony Tan might sound dramatic—but he has a point. If you’re not actively exploring AI, you’re preparing yourself (and your company) to become obsolete. The clock is ticking: Will you adapt, or will you become the adaptation?

What do you think?

Are you inspired or intimidated by Anthony Tan’s AI-driven future? Drop your thoughts below!

You may also like:

Grab and OpenAI’s Groundbreaking SEA AI Collaboration

AI Tsunami: Transforming Business Models in Asia

AI Ads Stir Up Conversations: The Future of Marketing in Asia

Ready more about Grab’s AI solutions by tapping here.

TL;DR – What You Need to Know in 30 Seconds

• PwC’s Agent OS orchestrates diverse AI agents into unified, scalable workflows, promising deployment up to 10x faster.
• Real-world cases already show efficiency boosts: 40% faster supply chains, 70% reduction in compliance tasks, and 30% quicker marketing launches.
• Designed for complex enterprise environments, it’s cloud-agnostic, multilingual, and actively deployed in leading global businesses, including PwC itself.

AI Agents are everywhere—but can they talk to each other?

PwC has unveiled its ambitious “Agent OS,” aiming to streamline AI orchestration at enterprise scale—promising workflows built and deployed 10 times faster. But is this platform truly the missing link enterprises need for their AI strategy?

Let’s dig in.

Enterprise AI sounds fantastic until you realise it often means managing a tangled web of different tools, platforms, and “intelligent” agents, each stubbornly refusing to play nice with each other. Companies regularly find themselves stuck between AI experiments and true enterprise-scale AI adoption because—ironically—these clever AI agents simply can’t collaborate.

Enter PwC’s new Agent OS, positioned as a kind of universal translator and orchestration conductor rolled into one. Imagine a central nervous system for enterprise AI, capable of seamlessly linking different agents and platforms into coherent workflows—no matter where the agents were developed or what tech stack they’re built on.

But is it all hype, or can PwC’s Agent OS genuinely unlock seamless, scalable enterprise AI?

What exactly is PwC’s Agent OS?

PwC’s Agent OS acts as a unified command centre, orchestrating a multitude of AI agents across popular enterprise platforms, including Anthropic, AWS, GitHub, Google Cloud, Microsoft Azure, OpenAI, Oracle, Salesforce, SAP, and Workday, to name just a few. It connects, coordinates, and scales AI agents—whether they’re custom-built, developed via third-party SDKs, or fine-tuned with proprietary data.

Think of it as the ultimate workflow builder, letting users—from AI specialists to your average non-tech-savvy manager—design sophisticated AI processes using intuitive drag-and-drop tools, natural language interfaces, and visual data-flow management.

Better yet, it’s cloud-agnostic, deploying effortlessly across AWS, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure, Salesforce, and even on-premises solutions.

Real-World Impact (not just theory)

Sceptical about fancy AI promises? Let’s look at some concrete use-cases PwC already claims are working in practice:

• Supply Chain: Imagine reducing your manufacturing firm’s supply chain delays by up to 40% through seamless integration of forecasting, procurement, and real-time logistics tracking agents from SAP, Oracle, and AWS, topped with PwC’s custom disruption detection agents.
• Marketing Operations: What if your retail marketing campaigns could launch twice as fast, with 30% higher conversion rates by orchestrating agents from OpenAI, Google Cloud, Salesforce, and Workday—all talking together in harmony?
• Compliance Automation: Picture a multinational bank automating regulatory workflows, drastically reducing manual reviews by 70%, thanks to agents seamlessly interpreting and aligning evolving regulatory policies via Anthropic and Microsoft Azure.

Who’s Already Benefiting?

PwC’s Agent OS isn’t just theoretical—real companies are already seeing transformative results:

• A tech company revamped its customer contact centre, reducing average call times by nearly 25%, slashing call transfers by 60%, and boosting customer satisfaction.
• A global hospitality firm automated brand standards management, achieving up to 94% reduction in manual review times.
• A healthcare giant applied AI agents to oncology workflows, streamlining clinical document processing to unlock actionable insights 50% faster, while simultaneously reducing administrative burdens by 30%.

And PwC themselves aren’t sitting idle: They’ve deployed over 250 internal AI agents, turbocharging productivity across tax, assurance, and advisory divisions—proving they’re ready to eat their own AI cooking.

Why PwC’s Agent OS Matters to Asia

In Asia, where enterprises are rapidly adopting AI to stay competitive (especially in dynamic markets like Singapore, India, and Indonesia), PwC’s Agent OS could offer a real edge. Asian enterprises grappling with complex multilingual data streams and diverse regional platforms may find a solution in the adaptive, multilingual capabilities of this system.

But it’s not just about tech. It’s about helping Asia’s leading enterprises quickly build, adapt, and scale AI-driven workflows to compete globally—accelerating innovation at a pace that keeps them ahead.

Could PwC’s Agent OS finally mean enterprises spend less time wrestling with AI tech—and more time reaping its benefits?

We’d love your take. Let us know in the comments below.

You may also like:

• How To Start Using AI Agents To Transform Your Business
• AI in Asia: Bridging the Risk Management Gap
• Microsoft’s AI Agents Set to Transform Asian Workplaces
• Read more about PWC agents on the official website by tapping here.