OpenAI

Since its founding, the Organization has undertaken a wide range of charitable initiatives, such as sponsoring a comprehensive basic income study, supporting economic impact research, and experimenting with education-centered programs like OpenAI Scholars. The Organizations efforts also included supporting a number of other public charities focused on technology, economic impact and justice, including the Stanford University Artificial Intelligence Index Fund, Black Girls Code, and the ACLU Foundation. More recently, the Organization forged partnerships with Duke University and other research organizations to sponsor additional AI impact studies. For example, the Organization facilitated a comprehensive task-based economic impact framework evaluating the potential economic impacts of large language models. The research found a potentially significant impact of artificial intelligence on the US economy. The research has been integrated into international policy discussions such as that by the OECD and the ILO, in addition to receiving broad coverage in the media. The Organization remains the governing body for all of OpenAI and holds a limited partnership interest in OpenAI LP. Each of the Organizations related organizations is legally bound to pursue the Organizations mission, such as following uncompromised principles of safety and broad benefit in its research and deployment efforts, unencumbered by profit incentives. Through its control of OpenAI, L.P., a capped-profit company formed to help rapidly scale the capital and talent needed to further its mission, and which the Organization governs and oversees all activities through its board, additional research milestones were accomplished in 2022 including the introduction of:1. OpenAI Residency program, which is designed to help bridge the knowledge gap for exceptional researchers and engineers in a variety of academic fields to gain the critical skills and knowledge to transition their careers into the AI & ML space. The Organization piloted this program by hiring, onboarding, and training a group of 14 residents from non-traditional backgrounds and fields of study for six months.2. Research Access Program, which supports researchers using OpenAIs resources to study areas related to the responsible deployment of AI and mitigating associated risks, as well as understanding the societal impact of AI systems. We provide early access to our models and tools to support their work across alignment, fairness & representation, safety evaluation development, model robustness, economic impact and other interdisciplinary research.3. DALLE 2, the second iteration of OpenAIs AI system that generates an original image from a description in natural language. The system was deployed iteratively as a creative tool with safety mitigations in place to prevent misuse. DALLE 2 includes a new technique that reduces bias and improves safety to produce results reflective of the worlds diverse population. Alongside the DALLE 2 deployment, the Organization also launched the OpenAI Artist Access program, which provides artists with subsidized and free access to OpenAIs tools.4. Embeddings, a new endpoint in the OpenAI Application Programming Interface (API) that makes it easy to perform natural language and code tasks like semantic search and classification.5. New and improved content moderation tooling to help developers protect their apps from misuse and detect undesired content using OpenAI classifiers. This safety tool is accessible to all developers free of charge.6. Whisper, an open-source neural net that approaches human level robustness and accuracy on English speech recognition.7. ChatGPT, a language model optimized for dialogue. It can answer follow-up questions and admit its mistakes. As a part of our approach to iterative deployment, a version is offered as a chatbot platform free of charges. The Organization actively monitors for misuse of its tools and continuously implements new safety features to ensure responsible use. These contributions to the advancement of AI technology, along with the continuous study on the broader societal impact of AI, are critical steps in moving the Organization towards the development of safe, effective, and productive general-purpose artificial intelligence that benefits all humanity.