Reimagining AI in Malaysia: An AOD x Elina Noor conversation

29 July 2026 - The Timber Space, Kuala Lumpur

On 29 July 2026, Architects of Diversity, together with independent researcher Elina Noor, convened Reimagining AI in Malaysia, a day-long conversation in Kuala Lumpur. Instead of the policymakers and industry figures who tend to dominate discussions about AI governance, the session gathered a cross-disciplinary group of participants across the arts, media, technology, climate activism and academia. 

Elina’s research brief, Sap and Silicon, provided the starting point. It traces a line from gutta percha sap, used to insulate the world’s first undersea cables in the 1850s, to today’s data centres. The main argument is that the extractive logic of the colonial era has not disappeared, but has simply taken new forms. Throughout the day, participants were asked to consider more fundamental questions, such as who benefits, who bears the costs and what kind of future does Malaysia actually want?

Discussion 1: What is AI actually for and who pays?

The morning brought a wide range of concerns to the surface. Some participants argued that Malaysia’s current AI buildout is being driven less by genuine social needs than by profit and short-term political incentives. As one participant put it, “when big-tech calls, you can’t say no.” The discussion often returned to themes of imperialism, colonialism and neocolonialism, with Malaysia described as a peripheral state that remains dependent on more powerful economies for AI technology, infrastructure and talent. 

From there, the discussion circled back to a more fundamental question. Participants asked what problem AI is actually solving. They suggested that there is an exposure gap between those who build the technology and people on the ground. As a result, the problems being addressed can sometimes feel imagined or abstract, while more immediate problems on the ground remain neglected. It was also noted how difficult it can be to bring sociologists and humanities scholars into these conversations in the first place. 

Accountability and anxiety were recurring themes. Participants felt that technology platforms, despite having enormous influence over how AI is developed and deployed, are not being held accountable effectively. At the same time, Malaysia appears to lack a clear national vision for what it actually wants from AI. Underlying much of the discussion was the fear of being “left behind”, and a sharper question of whether opting out is even possible. The group also explored collective action as a possible source of leverage, including the example of a coordinated technology blackout in Indonesia. 

Other topics ranged from the relationship between creativity and value, with the warning not to “confuse creativity with value”, to the decline of memorisation-heavy subjects such as History in schools. There were concerns about older people being scammed online, and about AI increasingly being used to answer religious questions without going through traditional authority channels. 

Discussion 2: Which futures remain open?

After a timeline-rewriting exercise revising missed opportunities and hidden actors across centuries, the room worked through six 2050 scenarios. A robot-helpers’ union threatens to strike, and Malaysia keeps its cultural values because it governs technology well. Bumiputera classification is abolished as race-based data trained on colonial categories is judged illegitimate. Malaysia’s first human-robot marriage ends in divorce, raising questions of legal personhood and AI representation. Malaysia becomes an irreplaceable chipmaker anchoring undersea cables through the Strait of Malacca. Public schools reach 100% AI literacy, opening discussions on public-interest AI versus commercial tools. The government subsidises AI tokens like petrol, leaving open whether that is empowerment or entrenched dependency.  

One comment captured the uncertainty running through the exercise. As one participant put it, “a boring future is preferable to an uncertain one.”

Final reflections

The closing discussion turned to sustainability and representation. The room was sceptical that data centres could ever be genuinely sustainable when the financial incentives remain focused on maximising profits. The deeper problem, participants argued, is an “unproductive capitalism” in which extraction appears cheap because the costs to land, water and the environment are left unpaid. The idea of building a national LLM also drew caution. Several participants worried that such a system could simply reproduce the worldview of the dominant group, repeating a much older colonial pattern in which whoever gets to document a culture also gets to define it. 

Elina pushed the conversation further by questioning the assumption that “more data” is always the answer. Some things, including silence, gesture and the weight of a gaze, resist being captured as data. To this end, choosing not to collect data can itself be a legitimate source of self-governance and resistance. 

AOD’s Jason Wee closed the day with a simple observation, that AI unsettles us partly because it has become “too human”. We fear it because we have created something that resembles ourselves, and urged a focus on how we coexist with it rather than dwelling on doom. 

As AOD continues its work to broaden the discussion on Malaysia’s own data-driven future, the day was a reminder that a grounded and nuanced conversation about technology is possible when a broader set of voices is meaningfully included. Malaysia’s current trajectory is not a foregone conclusion, and what the country chooses to build will depend on who is in the room to shape it. As Jo-Ann from Luminate put it in closing, “We don’t make the waves, but we can build the rafts to ride them” and this session, as she noted, was part of building those rafts. We extend our sincere appreciation to Elina, and to everyone who brought their insight and energy to the room.