OpenAI has just released GPT-5.5, and what stands out most is not just the intelligence leap but the combination of greater capability with latency matching the previous model. This resolves one of AI's most persistent tensions: more powerful models tend to be slower. OpenAI itself described GPT-5.5 as the most intuitive model it has ever released, capable of handling complex, multi-step tasks without requiring users to manage every step of the process. OpenAI What strikes me as most significant about this launch is the focus on genuine agency. This is not a model that simply answers better; it plans, uses tools, checks its own work, and navigates ambiguity on its own. OpenAI's Chief Research Officer noted that the model shows meaningful gains in scientific and technical research workflows, with concrete potential to accelerate breakthroughs in healthcare, including drug discovery. TechCrunch In a context like AfroSaúde's, where we sit at the intersection of technology, health, and equity, this kind of progress raises questions that go far beyond benchmark scores: who will have access to these capabilities? Will the populations most in need of mental health diagnostics and monitoring tools be included in this new phase? Intelligence is advancing quickly. Equitable access is not.