FAPESP launches ProCiêncIA program to fund long-horizon AI and HPC research

Brazil’s FAPESP ProCiêncIA program is a reminder that the race to AGI isn’t just about US and Chinese hyperscalers—it also depends on whether major emerging economies build serious research capacity. By putting up to five years of funding behind interdisciplinary networks that combine AI with high‑performance computing and domain problems like education, medicine, agribusiness and smart cities, São Paulo is effectively building local infrastructure for frontier‑adjacent work: better algorithms, better compute utilization, and richer application datasets. ([agencia.fapesp.br](https://agencia.fapesp.br/programa-estrategico-da-fapesp-em-ciencia-da-computacao-com-inteligencia-artificial-tem-sua-primeira-chamada/56870))

This kind of public investment tends to pay off indirectly for the global AGI effort. It produces trained researchers, open‑source tools and problem‑specific benchmarks that large labs can learn from, while also creating alternative centers of gravity so that advanced AI doesn’t become a purely North Atlantic and East Asian story. For local players in Latin America, access to stable, multi‑year funding for AI+HPC is a strong incentive to tackle harder problems instead of chasing short‑term app‑layer hype. If more regions follow this model, we may end up with a more geographically diverse set of contributors to core AI science, which can both accelerate progress and diversify perspectives on safety and governance.