IBM Corp. and Meta Platforms Inc. have announced the AI Alliance, bringing together companies, academic institutions, developers and scientists to promote the development of open AI models within an AI safety context. Other notable technology vendors that are founding members of the AI Alliance include AMD Inc., Cerebras, Dell Technologies Inc., Hugging Face, Intel Corp., Oracle Corp., Red Hat, ServiceNow Inc., Stability AI and Sony Group Corp. Notable absentees include AWS, Google, Microsoft Corp., NVIDIA Corp. and OpenAI. The group says it is open to new members in 2024 after a four- to six-week pause to establish some structure. The AI Alliance is also forming working groups, including one for safety and trust tooling.
The organization has six priorities, which are in summary as follows: to develop benchmarks, standards and tools to advance the production of multilingual, multi-modal, and science models that can help address society-wide challenges; to responsibly advance the ecosystem of open foundation models with diverse modalities; to foster a “vibrant” AI hardware accelerator ecosystem along with the accompanying software; to support global AI skills building; to develop educational content to inform the public and policymakers on “benefits, risks, solutions and precision regulation for AI”; and finally to launch initiatives, host events and showcase how Alliance members are using open technology in AI. This seems a broad set of priorities that IBM’s Anthony Annunziata, newly anointed head of AI open innovation, notes reflects the “diverse group of organizations” that make up the alliance.
The keyword used at the launch event for analysts was “open,” and it appears that the AI Alliance is designed to force other organizations to take a stance on whether they believe in open AI standards or not. Whether that necessarily means open source or not isn’t as clear — phrases used include “open science,” “open technologies,” “open innovation” and “open communities.” Some of the notable absentees mentioned are participating members in other standards and types of organizations that are also founding members of other bodies, such as the Linux Foundation and Partnership on AI.
The absence of some AI giants from the AI Alliance is perhaps as interesting as those organizations that have chosen to be members because it could indicate its purpose. The alliance may be designed as a counterweight to market leaders. The priority around influencing policymakers appears in this light to reflect a letter hosted by Mozilla on “open” AI development, which gained traction a few weeks ago. The letter criticized the idea that “tight and proprietary control of foundational AI models is the only path to protecting us from society-scale harm.” There is a perception that regulation, and the influence of those leading the generative AI pack, might lead to a concentrated and closed market.
Generative AI: From Hype to Value
Do you have your finger on the pulse of tech trends? Join the 451 Alliance for exclusive research content on industry-wide IT advancements. Do I qualify?
