Meta Launches PyTorch Foundation to Speed AI Research
Since 2016, when Mark Zuckerberg, CEO of Meta, joined the AI community to develop the PyTorch platform for AI technologies, PyTorch has become one of the top platforms for study and commercialization in the AI community, with thousands of collaborators and over 150,000 projects built on it.
Mark Zuckerberg stated today that the project would be transferred to the newly formed PyTorch Foundation, which will also be part of the nonprofit Linux Foundation, a technological collaboration whose primary objective is the joint advancement of open-source software.
The establishment of the PyTorch Foundation implies that decisions taken by a diverse range of members of the board will be transparent and open for many years to come. Because this decision will have long-term consequences in AI research, the governing board will be made up of officials from AMD, Metaverse, Microsoft Azure, Google Cloud, and Nvidia, with plans for expansion over time.
Zuckerberg Pledges to AI Research Driven by the Community
PyTorch was developed with an open-source, community-first attitude, and the shift to the foundation will not affect that. When developers and researchers open-source the software, people worldwide can work at their own pace, learn from each other’s advancements, and then give back to the AI community.
The collaborators to the framework will benefit in the future from the framework’s rigorous management, dynamic leadership, and additional capital given by the innovative PyTorch Foundation partners. Furthermore, the foundation will seek to follow four principles: openness, neutral branding, fairness, and establishing a powerful technical identity. One of its top goals will be to maintain a clear separation between PyTorch’s corporate and scientific management.
Meta Will Continue to Invest in AI Research
Meta will continue to make an investment in PyTorch as its principal platform for AI research and development. The shift will not affect PyTorch’s software, core projects, or developer operational strategies.
It has constantly worked to foster the community-driven growth that has powered PyTorch’s development, contributing numerous engineers to the foundation and assisting with technology enhancement and community engagement.
Its AI work is based on open science. The PyTorch Foundation believes that this approach will allow us to make and implement new solutions that will solve real-world demands while also answering basic concerns about the intelligence nature of AI. Furthemore, with the establishment of the PyTorch Foundation, the entire AI community is now equipped to move the field forward in a plethora of thrilling new directions.
