Every technological revolution creates new infrastructure. The greatest economic value, however, is often created by entrepreneurs who adapt that infrastructure to the realities of the markets. This is a story that has defined Latin America’s innovation journey for decades.
A handful of AI companies now account for an extraordinary share of global equity value. Hundreds of billions of dollars are flowing into chips, data centers, foundation models, and talent. Investor conversations have shifted from “will AI matter?” to “which AI company will win?”
Markets have already priced in an AI revolution. The real question is whether the economy has.
I recently listened to an Economist webinar on AI’s impact on financial markets. Afterwards, I found myself thinking about what three decades of building venture capital ecosystems across Latin America has actually taught me about technological change.
I was there when Latin America had no venture capital industry to speak of. I helped build it — first at the Inter-American Development Bank’s Multilateral Investment Fund, where we spent years convincing institutional investors that backing founders in Bogotá, São Paulo, and Mexico City was not charity but opportunity. We were early. We were sometimes wrong about timing. But the underlying conviction — that entrepreneurs closest to local problems build the most durable solutions — has never failed us.
That conviction shapes how I read the current AI moment.
The conversation dominating financial markets today is almost entirely about foundation models. OpenAI versus Anthropic versus Google versus whatever emerges next. These are genuinely interesting questions. But in thirty years of investing across Latin America, I have rarely seen lasting value created by whoever built the underlying infrastructure first. I have seen it created by the founders who understood their customers so deeply that the technology — whatever it was — became an unfair advantage in their hands.
This is not a contrarian position. It is what the history of technology in emerging markets consistently shows.
When mobile penetration accelerated across Latin America in the 2000s, the companies that ultimately mattered were not the carriers. They were the founders who understood that millions of people with a phone and no bank account represented an entirely new financial services market. When cloud computing arrived, the winners were not the data center operators. They were the SaaS founders who finally had the infrastructure to serve mid-market companies that legacy enterprise software had always ignored.
AI will follow the same pattern — but faster, and with higher stakes.
Across Latin America today, I see founders asking the right questions. Not “which model will win?” but “how do I use AI to do something for my customer that was simply impossible before?” How do you improve credit underwriting when formal income data barely exists? How do you optimize logistics routing across road networks that no algorithm was ever trained on? How do you help a mid-market manufacturer in Monterrey or Medellín finally get visibility into their own operations without a ten-year ERP implementation?
These are not small problems. Latin America’s productivity gap relative to developed markets is not primarily a capital problem or even a talent problem. It is an information and decision-making problem. Businesses operate with incomplete data, fragmented processes, and workflows that have never been digitized. Vertical AI — deployed by founders who have spent years learning about and learning the trust of these industries — has the potential to close that gap faster than any previous technology wave.
But there are other lessons from history.
Transformative technology and successful investment are not the same thing. I watched extraordinary optimism in 1999, in 2008, and again in 2021 lead otherwise disciplined investors toward shortcuts — chasing the headline, the valuation, the narrative, rather than the fundamentals. Every cycle, the story sounds different. The underlying mistake is always the same: confusing technological inevitability with investment certainty.
The productivity gains from AI will be real. The economic literature is increasingly clear that general-purpose technologies of this magnitude do eventually restructure how entire economies operate. But “eventually” has never been a straight line. Organizations need to redesign workflows. Workforces need new skills. Data — especially in markets where it has never been properly collected — needs to be structured before it can be useful. Culture, which moves slower than code, needs to shift.
What this means practically is that the most valuable applied AI companies in Latin America will be the ones that are quietly becoming indispensable to their customers — winning on retention, not headlines; on switching costs, not growth slides.
That is what durable venture returns have always looked like in this region: patient, specific, rooted in genuine customer understanding.
AI changes the magnitude of what is possible. It does not change what good investing requires.
The question worth asking today is not who will build the most powerful model. It is which founders, in which markets, already understand their customers well enough to make AI an unfair advantage.
I have spent thirty years backing Latin American founders before most of the world believed in them. AI does not change that bet. It gives even better odds.
Susana Garcia-Robles is Managing Partner at Capria Ventures.
This article was originally published on LinkedIn>
