Subscribe to The Informer for monthly expert analysis, and to Events for advance notice of visiting world leaders and distinguished guests.
You may unsubscribe from Lowy Institute newsletters at any time. For information on our privacy practices and how to unsubscribe, see our Privacy Policy.
The most-pressing world events explained by Lowy Institute experts and global contributors, in your inbox, every Thursday.
You may unsubscribe from The Interpreter at any time. For information on our privacy practices and how to unsubscribe, see our Privacy Policy.
Artificial intelligence, explained.
Meta founder Mark Zuckerberg’s earlier promises warrant scrutiny (Roberto Schmidt/AFP via Getty Images)
A Southeast Asian rejoinder to the superintelligence promise.
No, Mark Zuckerberg, the future is not (yet) for everyoneNo, Mark Zuckerberg, the future is not (yet) for everyone
Meta founder Mark Zuckerberg’s recent essay, “The Future Is for Everyone (Opens in new window)”, presents an ambitious vision of widely distributed superintelligence. Rather than concentrating advanced AI in a handful of technology companies and governments, individuals would have access to powerful personal AI systems that enhance their capabilities and help counterbalance concentrations of institutional power.
But the argument rests on two unexamined assumptions: that “everyone” is a population for whom access can simply be provisioned, and that broad access to superintelligence is inherently beneficial to humanity.
The appeal is understandable. Yet Zuckerberg’s earlier promises warrant scrutiny.
In 2015, he defended (Opens in new window) subsidised Facebook access in developing countries, arguing that everyone deserved access to tools and information supporting social and economic rights. With hindsight, we know that the main benefactor was Facebook (Meta) as it expanded its market into new territories.
In Myanmar, Facebook’s amplification and failure to moderate harmful content were linked to conditions in which violence could scale, with UN investigators documenting its role in spreading hate speech. In the Philippines, Facebook was likewise to amplify election-related disinformation during the 2022 elections.
The lesson is not that subsidised access caused these harms, but that rapid, poorly moderated platform expansions can create conditions in which harms could scale. That history warrants scepticism toward claims that providing AI for “everyone” is inherently beneficial.
The issue now is not whether superintelligence should be widely distributed, but who controls that distribution, under what safeguards, and to whose benefit.
Zuckerberg is not alone in making such appeals. Earlier technological innovations (Opens in new window) were often accompanied by grand promises of social transformation, except, interestingly and to some degree, the “godfathers” of AI (Opens in new window). Technology matters, but it does not automatically bring peace, equality, or social progress.
If we give Meta the benefit of the doubt and assume that it is acting benevolently, there is still a large and unjustified assumption. Zuckerberg argues that widely available superintelligence can empower individuals, stimulate innovation, and prevent excessive concentrations of technological power; namely, that access is non-problematic: provide affordable AI, connect it to a personal device, and individuals will benefit.
Southeast Asia demonstrates why this is not enough.
Access is a material and institutional condition, not simply a product feature. The region is highly uneven in technological capabilities. Singapore (Opens in new window) ranks among the world’s strongest governments for AI readiness, while capability gaps across ASEAN (Opens in new window) persist in connectivity, digital skills, and innovation networks. The infrastructure upon which AI relies – data centres, semiconductor supply chains, telecommunications infrastructure, and reliable electricity – is unevenly developed.
US–China semiconductor controls are also reshaping (Opens in new window) regional supply chains, while Southeast Asian countries are competing (Opens in new window) for data centre and AI investment. Access to superintelligence will depend not only on price, but on who controls the chips, energy, and infrastructure behind it.
Checking Facebook in Naypyidaw, Myanmar (AFP via Getty Images)
Zuckerberg also argues that distributing superintelligence to individuals could help prevent excessive institutional power. Unfortunately, distributing AI through a private corporation does not necessarily decentralise power. If Meta distributes powerful AI to individuals, it may expand users’ capabilities all while simultaneously strengthening Meta’s institutional power, market position, and control over the technology. The issue now is not whether superintelligence should be widely distributed, but who controls that distribution, under what safeguards, and to whose benefit. Widespread access to Meta’s AI is not necessarily the same as democratising superintelligence.
This concern is especially pronounced in Southeast Asia, where technology cannot be separated from diverse political systems, institutional arrangements, and levels of regulatory capacity. If Meta becomes a major intermediary through which individuals and governments access superintelligence, its role could extend well beyond providing technology to shaping how it is used and governed. The central question is whether distribution would genuinely disperse power or simply shift more power towards the corporation that controls the technology.
There is another dimension to making AI available to “everyone” – that is, whether the technology indeed understands everyone. Southeast Asia is linguistically and culturally diverse. AI can be widely accessible while still reflecting the assumptions, languages, and priorities of societies where its data and development are concentrated. Large language models such as Singapore’s SEA-LION (Opens in new window), developed around Southeast Asian languages and cultural contexts, show what more locally grounded AI development can look like. Distribution without representation risks making AI more accessible without making it more locally relevant.
Technology can amplify human capabilities, but it can also amplify existing inequalities and harms. The question, then, is whether everyone having access to superintelligence would actually make society better, and under what conditions it could make things worse.
For Southeast Asia, access is only the baseline. Who benefits, who bears the risks, and who has the power to shape how the technology is used will depend on infrastructure, skills, institutions, and whether AI reflects the region’s needs. That sets a much higher bar – one that cannot be decided and fulfilled by Silicon Valley alone.
About the authors
Karryl Kim Sagun Trajano
Karryl Kim Sagun Trajano is a Research Fellow for Future Issues and Technology (FIT) at the S. Rajaratnam School of International Studies (RSIS), Nanyang Technological University (NTU), Singapore.
Brendan Luyt
Brendan Luyt is Associate Professor of Information Studies at the Wee Kim Wee School of Communication and Information, Nanyang Technological University Singapore.
