Opportunities remain available for market entrants to build on AI models
Regulations on artificial intelligence (AI) do not necessarily inhibit smaller players in the market, where there are opportunities for use cases to be built on top of current major platforms.
Critics of regulatory policies, such as mandatory certification and licensing requirements, have argued that these rules will boost the foothold of large market players while rising entry barriers for startups looking to break into the market.
Also: Companies aren't spending big on AI. Here's why that cautious approach makes sense
Such regulation would crush innovation, said Andrew Ng, Stanford University professor and co-founder of Google Brain, on suggestions that AI could be made safer through mandatory licensing schemes. "There are definitely large tech companies that would rather not have to try to compete with open source [AI], so they're creating fear of AI leading to human extinction," Ng said in a recent interview with Australian Financial Review.
While he noted that the absence of policies is better than instilling bad ones, Ng pointed instead to the importance of having "thoughtful" regulation, such as the need for transparency from tech companies. This action could have helped prevent the harm these companies created with social media and would navigate the industry away from a similar result with AI, he said.
AI regulations, though, do not necessarily inhibit startups and market entrants, said Florian Hoppe, partner and head of vector in Asia-Pacific at Bain & Company, in response to ZDNET's question on the impact of legislation on AI innovation.
Also: 4 ways to detect generative AI hype from reality
Large language models (LLMs), for instance, are costly to build and smaller players typically will lack the resources to develop their own. However, there are opportunities for new use cases to built on top of existing LLMs, such as specialized or domain-specific AI applications and models, Hoppe said.
Startups will be able to develop such products without the constraints of having to build their own LLMs, he noted, adding that regulations play a necessary role in mitigating AI risks.
Conversations have also been healthy between governments and industry players on how the regulatory framework for AI should evolve moving forward, added Sapna Chadha, Google's Southeast Asia vice president. This situation is true for the region, which is technology- and digital-forward, setting the right path ahead for Southeast Asia markets to strike a good balance between the need for regulation and the requirement to drive market innovation, Chadha said.
A conducive environment will be essential to ensure that AI can bring about economic and business benefits, while safeguarding against potential risks, such as data bias, said Fock Wai Hoong, Southeast Asia head for Singapore's state-owned investment firm, Temasek Holdings.
Driving Southeast Asia's digital economy revenue toward $100B
In fact, data infrastructure and regulation are among key enablers that will push the region to become a sustainable digital economy, according to the latest e-Conomy SEA (Southeast Asia) report released by Google, Temasek, and Bain & Company.
Also: AI at the edge: Fast times ahead for 5G and the Internet of Things
Investments in digital and physical infrastructures and economic development plans will better enable digital organizations to expand services to areas outside metro cities in the region, where demand for digital products and services is growing, the report noted. These investments, if done right, can drive digital adoption and reduce the cost to serve.
According to this year's e-Conomy report, the region has weathered global macroeconomic headwinds better than other regions, with GDP growth at above 4% and consumer confidence showing a rebound in the second half of 2023, after falling to lower levels in the first half.
The Southeast Asian digital economy is projected to hit $100 billion in revenue this year, clocking 27% in compound annual growth rate since 2021, and growing 1.7 times as fast as gross merchandise value (GMV). E-commerce, travel, transport, and media will account for $70 billion in revenue, the report estimates.
Also: As developers learn the ins and outs of generative AI, non-developers will follow
GMV is expected to expand 11% year-on-year to $218 billion in 2023, with travel and transport on track to exceed pre-pandemic heights next year.
E-commerce also remains on a growth path this year, increasing 22% in revenue year on year to hit $28 billion. GMV in the sector is projected to climb to $139 billion in 2023, before hitting $186 billion in 2025 on a 16% growth rate.
To further sustain digital growth in the region, the report points to the need for digital businesses to focus on monetization and establish a path to profitability. In addition, digital inclusion remains crucial, and governments in the region must continue to invest in infrastructure building and to plug connectivity gaps in rural areas. This will ensure digital services are accessible in geographies where there is growing consumer demand.
The report also calls for the development and harmonization of policies and agreements, such as trade and data governance agreements, across Asean and to ease cross-border data flow and digital economy activities.
A policy framework for responsible AI development, for instance, can encompass "balanced legal frameworks" for AI innovation, with privacy laws to safeguard personal data and enable trusted cross-border data flows.
Also: Organizations are fighting for the ethical adoption of AI. Here's how you can help
AI governance frameworks should also be interoperable across the region and globally, with the development of common standards and shared best practices to ensure AI technologies are developed and adopted responsibly.
Southeast Asia is showing resilient growth, but the region is fragmented with diverse markets and siloed policy frameworks, Fock said. He urged for focus on establishing unified and multilateral agreements, such as digital economy agreements, and pulling together disparate efforts between the individual markets.
A "single digital Southeast Asian" market, where there is interoperability and seamless connectivity, could be the model to drive growth in the region, he said. Singapore, for instance, already has digital economy agreements with nations such as France, New Zealand, and the UK, and can look to make similar pacts with its peers in Southeast Asia.
Asean member states in September said that they are working to establish protocols that will ease cross-border digital trade and help address emerging trends, such as AI. Targeted to be completed by 2025, the Asean Digital Economic Framework Agreement will improve digital rules across key areas, including digital trade, cybersecurity, payments, and data. The framework aims to facilitate seamless cross-border online trade and make it easier to do business within the region.
Asean has also championed its unified efforts in cybersecurity and pledged to drive further collaboration among member states, including plans to adopt common standards and best practices. To date, Asean is the only regional organization to have subscribed, in principle, to the United Nations' 11 voluntary, non-binding norms of responsible state behavior in cyberspace.