We are focused on research and development of leading-edge generative AI models like GPT-3. These models are capable of sophisticated text generation and have many potential applications.
However, the unprecedented power of these models raises concerns about potential misuse or lack of alignment with human values. We aim to proactively address these risks.
A core part of our work is thus on AI safety - ensuring powerful generative models behave in beneficial ways aligned with ethics and human values.
We want to make these models open source and freely accessible to the general public, not just large tech firms. This increases transparency and accountability.
By developing high-capability AI but with openness and care for societal impact, we hope to distribute benefits widely in a responsible manner.
Our goals combine cutting-edge AI capabilities with ethics and concern for the common good. We aim to steer generative AI in a direction that uplifts humanity.
Making powerful AI accessible, but in a thoughtful way, allows more people to gain an advantage from AI while managing risks responsibly.
In essence, our work fuses inclusion, transparency, and ethics with research on the most advanced generative models. The aim is democratizing AI safely.