Our mission is to build human-like general intelligence and make it safely accessible in order to foster a more abundant, equitable, and unconstrained society.
We take a first-principles approach, starting from simple self-supervised architectures and evolving them to tackle human developmental milestones of increasing complexity.
We believe human-like artificial general intelligence can enable a society of greater abundance and opportunity by reducing bottlenecks to progress, expanding each person's potential, and giving insight into the human mind.
If you're a fearless explorer interested in a principled approach to building safe and benevolent artificial general intelligence, come work with us!
Think good means earnestly—with extraordinary curiosity and without ego—trying to understand what is actually true. For example: clarifying when a concept feels fuzzy, interrogating assumptions, steelmanning counterarguments, and digging into inconsistencies.
Be kind means starting from a place of psychological safety—having our teammates' best interests at heart while trusting that they have ours—and embracing radical candor. Being kind not only brings us closer to each other but also embodies our philosophy for how AGI can positively impact all of humanity.
Having fun is a critical part of great work; it stimulates creativity and resets stress. We devote every Friday afternoon to hanging out together in SF, doing everything from playing D&D and trying new coffee shops to building funny robots.
Growth is like compound interest: where you end up depends much more on your rate than on your starting point! Grow together takes this seriously. We have a weekly paper club and we read textbooks as a group. In addition to learning together, we each take the initiative to identify areas for personal growth and support each other's growth goals.