"Move Over ChatGPT" Meta-IBM Form AI 'Open Source' Alliance

Intel, AMD, NASA Participate
Joint Technology Development and Risk Response

Meta Platforms, IBM, Intel, AMD, and others have formed an alliance to build an artificial intelligence (AI) ecosystem through an open-source strategy by releasing large language models (LLMs) for free. This move contrasts with major AI companies like OpenAI, Microsoft (MS), and Google, which have maintained confidentiality regarding their technologies, and aims to contribute to the creation of a new AI ecosystem.

[Image source=Reuters Yonhap News]

On the 4th (local time), IBM and Meta announced in a press release that 57 companies and organizations have come together to establish the 'AI Alliance.' This alliance includes major tech companies such as IBM, Meta, Intel, AMD, Oracle, Dell Technologies, Sony Group, as well as AI companies like SakanaAI, Silo AI, and Stability AI. Universities including Yale University, Cornell University, Dartmouth College, Hebrew University of Jerusalem, Imperial College London, University of Tokyo, and Keio University, along with U.S. government agencies such as NASA and the National Science Foundation (NSF), have also joined.

The AI Alliance believes it is important to develop AI technology in an open manner. They hold that AI technology should be developed publicly with the participation of many stakeholders so that more people can benefit, innovative products can be created, and various risks that may arise can be identified and mitigated in advance through diverse methods.

Unlike OpenAI, Microsoft (MS), Google, and Amazon, which maintain confidentiality and do not publicly disclose their AI technology development, Meta has been developing AI technology using an open-source approach since February. This is interpreted as an effort to unite companies with similar needs, related startups, research organizations, and government agencies to strengthen ecosystem building.

Earlier, Meta released its own LLM called 'Llama' in February, making the related technology public. Initially, commercial use was not permitted, but five months later, in July, Meta released Llama 2, providing it as open source for free use in research and commercial applications.

Mark Zuckerberg, CEO of Meta, said at the time of Llama 2’s release, "I believe that the more open the ecosystem is, the more progress will be possible," adding, "That is exactly why we released the open-source model Llama 2." He emphasized, "When software is open, more people can carefully examine it to identify and fix potential issues."

After observing Meta’s approach, IBM contacted Meta in August and proposed the establishment of the AI Alliance, which initiated discussions.

Dario Gil, Senior Vice President at IBM, told The Wall Street Journal (WSJ), "Honestly, I have been dissatisfied with the overall AI discussions over the past year," explaining, "I felt that the diversity of the ecosystem created by the AI moment was not sufficiently reflected." He added, "It would be a big mistake to think that the future of AI will be decided by two, three, or five institutions," and said, "I hope the world will have confidence to bet on open innovation through the AI Alliance."

The AI Alliance is expected to soon release tools for AI safety and model verification.

However, foreign media are paying attention to the fact that the AI Alliance emerged one year after the launch of ChatGPT. The WSJ pointed out that most companies participating in this alliance have their own AI products but have not received as much attention as OpenAI and MS. It is suggested that companies trailing in the AI leadership race chose the AI Alliance as a means to create a competitive structure together before it is too late.

There are still critical views regarding the development of AI technology through open source. Given the lack of AI-related regulations and the many challenges that the technology itself still faces, there are concerns that open source could lead to greater damage. It is feared that the use of open source could cause explosive increases in spam, financial fraud, and misinformation.

This content was produced with the assistance of AI translation services.