A new frontier AI lab, humans&, has emerged from stealth, announcing a seed round backed by Jeff Bezos, NVIDIA, and Google Ventures (GV). The startup is founded by a supergroup of researchers and builders who previously led major efforts at xAI, Anthropic, Google DeepMind, and OpenAI, and aims to fundamentally pivot the industry’s focus away from autonomous agents toward truly human-centered AI.
The core premise of humans& challenges the current trajectory of AI development. While large language models are rapidly improving their ability to reason, code, and act independently, the humans& team argues that real-world progress stems from human connection, trust, and collaboration. They state that the next chapter of AI should begin by centering around people and their relationships, serving as a "deeper connective tissue that strengthens organizations and communities."
This shift requires significant technical retooling. The company notes that building collaborative AI demands innovations in long-horizon and multi-agent reinforcement learning, advanced memory systems, and nuanced user understanding—areas where current models often struggle.
The All-Star Pedigree and the Funding War Chest
The launch of humans& is notable less for its mission statement—many startups claim to be "human-centric"—and more for the pedigree of its founding team. The roster includes veterans who have shaped modern AI, having collectively shipped models and products used by billions. Their combined experience spans seminal work in reasoning, behavioral training, and alignment across nearly every major frontier lab.
The immediate financial backing signals serious intent to compete at the highest level. The seed round was led by SV Angel and co-founder Georges Harik, but the inclusion of strategic investors like NVIDIA, Jeff Bezos, and GV (Google Ventures) ensures the lab has the resources and hardware access necessary to train models at scale. Emerson Collective and Forerunner also participated.
By tightly integrating science and product development, humans& intends to drive a new paradigm where AI is measured not by its individual autonomy, but by its ability to facilitate human understanding. The lab has already committed to contributing back to open source and academic research, indicating a long-term strategy to influence the broader direction of human-centered AI development.
