Today’s article will explore:

• Southeast Asia's AI potential;

• What are the investors looking for?

• Emerging AI trends that are gaining traction; and

• Hurdles in AI's success in the SEA region.

Potential Value of AI Applications

Companies like Amazon, Google, Microsoft, and Meta collectively spent $177 billion in capital expenditures, half allocated to infrastructure like AI data centers. 

Key point: Most current AI investments are made 10-15 years ahead of expected revenue.

Presently, the AI industry is dominated by infrastructure-level players (Nvidia chips, AI data centers), with cloud providers offering ‘GPUs as a Service.’ 

However, the maximum value from AI applications will materialize as infrastructure advances to ensure accessibility and reliability. While AI applications hold tremendous potential, their full impact will be realized when the supporting infrastructure is mature enough to handle widespread deployment. It's not a matter of one over the other—infrastructure and applications will evolve together, driving significant value creation as enterprises and consumers become ready to adopt and benefit from these solutions.

AI investors (VCs and tech companies) are increasingly bullish on infrastructure and model development. Much of the success in AI applications, like ChatGPT, Claude, video-01, and talkie.ai, directly results from investments in these foundational models and infrastructure.

While these applications are making headlines, new AI applications are not raising funds as easily as the major players. They are still very early (almost in infancy) and generate around $20 billion in revenue.

So what are some of the AI trends that we are going to see going forward, especially in the Southeast Asia market?

New AI Application Trends in Southeast Asia:

There’s been a surge in new AI applications emerging from Asian countries, especially in the B2C space. Companies are increasingly experimenting with AI, and B2C apps dominate because they allow for faster iteration cycles and cost-effective hypothesis testing. This makes it easier for businesses to refine and scale AI solutions quickly, leading to a higher concentration of successful consumer-facing applications.

Some of the cities with consumer AI apps ranked among the top 100 globally:

• Gurugram

• Kuala Lumpur

• Islamabad

• Singapore

• Seoul

• Hanoi, 

• Ho Chi Minh City

Six months ago, most web traffic in AI was centered around LLM-based assistants like ChatGPT. However, recent analysis reveals a few new, rapidly growing categories:

1. Music: The most prominent example is ‘Suno.’ The tool generates original songs in the browser from a text prompt, complete with lyrics, in various styles.

2. Productivity: seven out of the top 100 ranked AI apps (according to a report by a16z) are all productivity based:

   • Liner

   • Eightify

   • Find

   • MaxAI

   • Blackbox AI

   • Otter.ai

   • ChatPDF.

   These categories are gaining traction as users look for solutions that can seamlessly integrate into their daily workflows, automating tasks and boosting efficiency. In a region like Southeast Asia, where speed and adaptability are crucial, productivity apps are essential for managing personal and professional tasks.

But the most intriguing trend is:

3. The rise of AI companionship: Six months ago, there were just two AI companion companies (in the top 100), but today, there are eight on web platforms and two on mobile.

   Character AI, a prominent example, reports that users spend an average of two hours per day interacting with its AI companion. This signals a shift from AI being merely a tool to becoming a source of emotional and social engagement.

   

   Top consumer AI mobile apps are now heavily pursuing this companionship category, reflecting the demand for AI that offers both functional and personal connections.

   As this trend continues, AI companionship will evolve into more sophisticated and immersive experiences driven by the need for more personalized, emotionally responsive AI solutions.

   This shift represents a broader move toward AI applications that deeply integrate into everyday life, not just solving problems but also enhancing personal well-being and engagement across diverse user needs.

Which AI Companies Will Lead the Southeast Asia Region? 

Southeast Asia presents a unique challenge for building a scalable AI application because of its cultural and linguistic diversity. Unlike other regions worldwide, SEA comprises countries with drastically different languages, traditions, mindsets, and consumer behaviors.

That makes it extremely difficult for a single AI application to achieve dominance across the region. For instance, an AI solution that resonates in Singapore might not have the same impact in Indonesia or Vietnam. Success in Southeast Asia often depends on localizing AI applications to fit each country's consumer behavior differences. Companies that tailor their solutions to meet local needs while maintaining a universal core will have a better chance of scaling across the region.

As AI matures in SEA, applications that cater to regional adaptability will stand out.

Fintech applications can be a perfect example because transactions are almost universal. The only difference is how users are acquired and retained. But that's not a technology problem.

By focusing on challenges affecting users across borders—such as improving workflow efficiency, enhancing quality of life, or automating mundane tasks—AI companies can create products with broad appeal while still accounting for the region’s diversity.

This strategic approach will allow AI startups/companies to penetrate and thrive in Southeast Asia's complex and varied markets.

• AI potential in Southeast Asia is enormous, with projected economic gains of $950 billion by 2030.

• AI investments to date are largely infrastructure-focused and made with a 10-15 year horizon, positioning for future application development.

• Emerging AI trends like productivity tools, music, and AI companionship are gaining traction, especially in SEA.

• Applications that focus on solving universal problems that cut across regional diversity will provide the maximum AI value in SEA.