The talent disperses, most works in isolation.
Latin America stands at a familiar crossroads: rich in human potential, yet unable to translate that potential into collective power. Nayat Sánchez-Pi, research director at Inria Chile, argues that the region's struggle with artificial intelligence is not a talent problem but a structural one — the absence of institutions capable of organizing scattered brilliance into lasting scientific momentum. As AI reshapes the global order, the question for Latin America is no longer whether to participate, but whether it can build the coordinating frameworks that turn individual excellence into regional sovereignty.
- Latin America produces world-class AI researchers, yet the region watches from the margins as others set the global agenda — talent alone is not enough when there are no institutions to hold it together.
- The arrival of AI risks amplifying an old wound: researchers working in isolation on fashionable problems, their breakthroughs dissolving before they can compound into something larger.
- Sánchez-Pi calls for a fundamental reframing — treating AI not as a computer science specialty but as foundational infrastructure, like electricity, capable of accelerating work across health, biology, and climate science.
- Fragmented funding and the absence of long-term institutional vision threaten to produce islands of innovation surrounded by disconnection, deepening inequality rather than spreading technological gains.
- The region's own assets — Brazilian biodiversity, Chilean astronomical infrastructure, unique ecosystems — offer a path to an original AI agenda, one built for Latin American realities rather than borrowed from the Global North.
Latin America has no shortage of artificial intelligence talent. Yet the region remains a supporting player on the global stage, never quite the lead. Nayat Sánchez-Pi, who directs research operations at Inria Chile, has a blunt diagnosis: the problem is not the people. It is the absence of structures that would allow those people to matter.
Speaking at an event on the future of science hosted by Hospital Albert Einstein, Sánchez-Pi frames the challenge as a mismatch between individual capability and institutional capacity. The region produces researchers of genuine quality but cannot organize them into something larger than the sum of their parts. Talent disperses, collaborations remain thin, and what emerges is a scatter of individual efforts rather than a coherent scientific movement. She calls this the paradox of solitude — and warns that AI, rather than solving it, risks deepening the pattern.
The path forward, she argues, begins with a shift in how the region thinks about AI itself. The question is no longer whether to adopt it, but how to deploy it strategically. Sánchez-Pi proposes treating AI as foundational infrastructure — the way a society treats electricity or water — so that it can serve as a platform across disciplines, accelerating work in health, biology, and climate monitoring alike. This requires institutional commitment: long-term vision, coordinated public investment, and a deliberate effort to ensure that advances spread beyond elite centers.
What makes this vision credible, she suggests, is what Latin America already possesses. Brazilian biodiversity, Chilean astronomical facilities, ecosystems found nowhere else — these are assets the Global North cannot replicate. Rather than copying models built for different contexts, the region could build its own AI agenda, using the technology to answer scientific questions that only it can ask. Whether the institutions willing to pursue that vision will emerge remains, for now, the open question.
Latin America has no shortage of artificial intelligence talent. Walk into any research lab from São Paulo to Santiago and you'll find capable scientists working on cutting-edge problems. Yet the region remains a supporting player on the global stage, never quite the lead. Nayat Sánchez-Pi, who directs research operations at Inria Chile, has spent considerable time thinking about why. Her conclusion is blunt: the problem is not the people. It is the absence of the structures that would allow those people to matter.
Sánchez-Pi, speaking recently at an event on the future of science hosted by Hospital Albert Einstein, frames the regional challenge as a fundamental mismatch between individual capability and institutional capacity. Latin America, she argues, has landed on the global technology map. But it has not learned to lead from there. The region produces researchers of genuine quality, yet it cannot seem to hold onto them or organize them into something larger than the sum of their parts. The talent disperses—some stays, some leaves, most works in isolation. What emerges is not a coherent scientific movement but a scatter of individual efforts, each pursuing popular topics with minimal collaboration.
The arrival of artificial intelligence has made this problem more visible, not less. AI has the potential to accelerate scientific careers both individually and collectively. But in Latin America, the technology threatens to deepen an existing pattern: researchers working alone on fashionable problems, their breakthroughs never quite connecting into something systemic. Sánchez-Pi calls this the paradox of solitude. The region risks using AI to do what it has always done—produce isolated excellence that never quite compounds into regional power.
The path forward, she argues, requires a shift in how the region thinks about both AI and itself. The debate has moved past whether Latin America should adopt artificial intelligence. That question is settled. The real question is how to deploy it strategically. Sánchez-Pi proposes treating AI not as a specialized corner of computer science but as foundational infrastructure—the way a region might treat electricity or water systems. When AI becomes infrastructure, it can serve as a platform for work across disciplines: health, biology, climate monitoring. Scientists in different fields can build on it, accelerate their work through it, reach results faster and with greater precision.
This infrastructural approach requires institutional commitment. It demands long-term vision, not the fragmented investments that currently characterize regional funding. It requires public institutions willing to coordinate, to set standards, to ensure that advances do not remain trapped in elite centers but spread across society. Without this coordination, the region risks creating islands of innovation—pockets of excellence surrounded by disconnection. The goal instead should be democratized access, ensuring that technological gains reach different sectors and different communities.
What makes this vision possible, Sánchez-Pi suggests, is the region's own assets. Latin America possesses unique natural resources and scientific infrastructure that the Global North does not: Brazilian biodiversity, Chilean astronomical facilities, ecosystems and environments found nowhere else. Rather than simply copying models developed in Europe or North America—models built for different contexts, different genetic diversity, different social realities—Latin America could build its own agenda. It could use AI to unlock the particular scientific questions that only this region can answer. That would require institutions willing to think long-term, to coordinate talent, to treat technology as a tool for regional autonomy rather than regional dependence. Whether those institutions will emerge remains the open question.
Notable Quotes
The region is on the global technology map, but it does not lead from there— Nayat Sánchez-Pi, director of Inria Chile
What could truly position the region is not individual work, but research institutions and an organizing approach that creates positive collective effects— Nayat Sánchez-Pi
The Hearth Conversation Another angle on the story
When you say Latin America has talent but lacks institutions, what exactly breaks down? Is it money, or something else?
It's partly money, but more fundamentally it's coordination. You can have brilliant researchers scattered across a continent, each working alone on their own problem. They never build on each other's work. There's no shared infrastructure, no long-term commitment. The talent leaks away or stays isolated.
So the brain drain isn't just people leaving—it's also people staying but not connecting?
Exactly. A researcher might stay in Brazil, but if there's no institutional framework supporting their work, no way to collaborate regionally, they're still isolated. They might as well be working alone.
You mentioned treating AI as infrastructure. What does that actually look like in practice?
It means public institutions building AI systems that scientists across different fields can use—biologists, climate researchers, health specialists. Right now AI is treated as a computer science specialty. But if it becomes a shared tool, a foundation, then everyone benefits from it.
And the regional assets—biodiversity, astronomy—how do those change the equation?
They're the reason Latin America doesn't have to copy the North. Those assets are unique. If the region builds institutions around them, using AI to study problems only Latin America can study, then you're not following someone else's agenda. You're building your own.
What happens if the region doesn't make this shift?
The talent keeps dispersing. The breakthroughs stay isolated. Latin America remains a place where smart people work, but never the place where the future gets decided.