Meta, IBM Look to Drive Open Innovation With AI Alliance

Dozens of universities and technology companies across the world have banded together to form an alliance that intends to support open innovation in artificial intelligence.

See Also: Live Webinar | All the Ways the Internet is Surveilling You

Meta and IBM are leading the group, called AI Alliance, which includes major players such as Dell, Intel and NASA, as well as a dozen universities and research institutes. The group seeks to address education, research, development, deployment and governance aspects of AI. The group does not include OpenAI.

The alliance seeks to allow its members to collaborate and share information that will help hasten innovation and develop solutions that fit the needs of researchers, developers and adopters while identifying specific risks and mitigating them before putting a product into the world, Meta and IBM said in identical statements.

The U.S. National Science Foundation, which is a member of the alliance through the National AI Research Resource, said it will make the resources and tools the group develops available to the research community to boost innovation in areas of societal and economic importance.

"It is better when AI is developed openly. More people can access the benefits, build innovative products and work on safety," said Nick Clegg, president of global affairs at Meta.

The AI Alliance will form working groups to tackle specific areas of innovation and will be governed by a board and technical oversight committee that will determine project standards and guidelines. It also plans to partner with existing initiatives from governments, nonprofits and civil society organizations whose work aligns with the group's goals.

Kevin Murphy, chief science data officer at NASA, said open innovation is "all but essential" to allow equitable access and collaboration around AI. The companies said that the alliance will support advancement of open foundation models that have "diverse modalities, including highly capable multilingual, multimodal and science models that can help address societywide challenges in climate, education and beyond." The alliance is expected to hold space for initiatives that showcase its members using open technology in AI "responsibly and for good."

"Open source is the backbone of all leading artificial intelligence software. With open source, the entire community comes together to collaborate on solving the toughest problems, the most effective solutions rise to the top, and everyone benefits," said Jeremy Howard, founding researcher at AI Alliance member Fast.ai.

The group also seeks to develop benchmarks and evaluation standards, including setting up a catalog of vetted safety, security and trust tools for the responsible global development of AI. It also plans to enable an AI hardware accelerator ecosystem, support skills building and exploratory research, and develop educational content for public discourse and to inform policymakers about the benefits, risks, solutions and precision regulation for AI.

Innovations in artificial intelligence offer both "the promise of serving the common good and the threat of undermining it," said John I. Jenkins, president at the University of Notre Dame, which is a member of the alliance. He said that it is "critical" to determine solutions to ethical implications while researching technological advancements.