Meta, a global social and technology giant, announced on its official website that it has formed an AI Alliance with 57 technology and academic research institutions including Oracle, Intel, AMD, IBM, Sony, and Dell.
Many of these 57 organizations areLarge ModelOpen SourceFieldLeaders, for example, Meta, the open source originator of ChatGPT-like models;
Stability AI, the most well-known open source platform in the field of human graphics;FamousHugging Face, one of the large model open source sharing platforms.
so,This AI alliance is actually more like an "open source Avengers alliance" with a lot of talent, there are AMD and Intel that make chips; Meta and Stability AI that make underlying large models;
Oracle and IBM are engaged in the software application layer; Cornell University, University of Tokyo, Yale, Imperial College London and other famous universities are engaged in academic research, forming a perfect closed loop from research and development to application.
Today's big model world is divided into two camps: open source and closed source.The AI alliance led by Meta is obviously intended to challenge ChatGPT, Midjourney and Gen-2 in the closed-source field..
These three companies are the three big losers in the fields of text, pictures and videos, and are very successful in terms of product influence and commercialization.
However, since it has not been open source, many companies that want to use it can only spend a lot of money, which will lead to a resource monopoly in the long run.
In fact, this can also be seen from the complete list of members. The three giants in the field of large models today - Microsoft, Nvidia, and OpenAI are not on it.
The six goals of the AI Alliance
Among these six major goals, the main emphasis is on building an open source big model ecosystem, including a complete set of processes from research, evaluation, hardware, security, public participation, etc.Simply put, it is about grabbing businesses, grabbing users, grabbing resources, and grabbing the right to speak.
1) Develop and deploy benchmarks and evaluation standards, tools, and other resources to enable responsible development and use of AI globally, including the creation of a catalog of vetted safety, security, and trust tools. Support the promotion and dissemination of these tools to the developer community for large-scale model and application development.
2) Responsibly advance an ecosystem of open source models in different modalities (text, audio, images, video, etc.), including high-capacity multilingual, multimodal, and scientific models, to help address societal challenges in areas such as climate and education.
3) Foster a vibrant AI hardware accelerator ecosystem by promoting the contribution and adoption of important software technologies.
4) Support global AI skills building and exploratory research. Engage the academic community to support researchers and students to learn and contribute to important AI big model and tool research projects.
5) Develop educational content and resources to educate the public about the benefits, risks, solutions, and safety regulations of AI.
6) Launch initiatives to encourage open AI development in a safe and beneficial way, and host events to explore AI use cases and showcase how Alliance members can use AI technologies responsibly and for public good.
How the Alliance Works
The AI Alliance will initially establish member-driven working groups in all of the major thematic areas listed above.
The Alliance will also establish a Management Committee and a Technical Oversight Committee to specifically advance work in the above project areas and develop overall project standards and guidelines.
In addition to bringing together leading developers, scientists, scholars, students and business leaders in the field of AIleaderThe AI Alliance will also plan to work with important existing initiatives in government, nonprofit, and civil society organizations that are engaging in valuable collaboration in the field of AI.
For more details: https://thealliance.ai/
Full Membership List
