Technology giants that include IBM Corp. and its Red Hat subsidiary, Meta Platforms Inc., Intel Corp., Advanced Micro Devices Inc. and ServiceNow Inc. said today they’re joining forces with more than 40 other organizations across business, academia, government and advocacy to create a new global initiative on the development of artificial intelligence.

The new group is called the AI Alliance and it intends to focus on encouraging open innovation and open science in AI development. The idea behind the initiative is to help shape the evolution of AI in such a way that it can better reflect the needs and complexities of the world’s societies.

In an announcement, the founding members of the AI Alliance say the rapid pace of advancement in AI is creating new opportunities everywhere that can improve the way people work, live, learn and interact with one another. To take advantage of this, open and transparent innovation is required to empower the broadest possible spectrum of AI researchers, builders and adopters, in order to provide them with all of the tools they need to harness AI in a way that prioritizes safety, diversity and economic opportunity.

The alliance avoids criticizing the highly secretive approach to building AI systems that’s employed by industry leaders such as OpenAI, and instead simply says that more collaboration and information sharing can help the AI community to innovate much more quickly, in a more inclusive way. This open approach to AI development will also help to identify the specific risks posed by AI development and help to mitigate them before dangerous systems are released into the world, the alliance said.

“The AI Alliance is focused on fostering an open community and enabling developers and researchers to accelerate responsible innovation in AI while ensuring scientific rigor, trust, safety, security, diversity and economic competitiveness,” it said in a statement.

That said, it’s notable that the alliance doesn’t include AI and cloud leaders such as OpenAI, Google LLC, Amazon Web Services Inc., Microsoft Corp. or even Nvidia Corp.,, so it remains to be seen how effective it will be. Rob Strechay, an analyst with SiliconANGLE Media’s theCUBE Research arm, said that if governments view the AI Alliance as an important organization to work with, “this could be a problem for those organizations that don’t want to at least participate.”

TheCUBE Research Chief Analyst Dave Vellante noted that those companies’ absence isn’t completely surprising, given their leadership in AI mindshare. But he said Amazon would be a natural to lead an alliance based on opening up AI to more players beyond the current leaders, given the cloud giant’s focus on providing a wide range of AI models in its offerings.

AI Alliance members will focus on pooling their resources and knowledge to address the safety concerns of AI development and create an open platform for sharing and building systems that fit the needs of researchers, developers and adopters.

To do this, the alliance is planning to lead multiple collaborative research projects. For instance, it aims to develop and deploy benchmarks and evaluation standards, plus tools and other resources that can ensure the responsible development and use of AI systems globally. The organization also wants to advance the growing ecosystem of open-source foundation models and ensure everyone can access powerful multilingual, multimodal and science-based models that can tackle global challenges in climate, education and more.

Another goal of the alliance is to foster a vibrant AI hardware accelerator ecosystem through the development of essential enabling software technologies. Meanwhile, the academic community will be encouraged to support AI researchers and students, helping them to learn and contribute to the development of open models and tooling. There will also be public outreach, with the alliance planning to develop educational resources that will better inform public discourse and lawmakers on AI’s benefits, risks, solutions and regulatory needs.

“The progress we continue to witness in AI is a testament to open innovation and collaboration across communities of creators, scientists, academics and business leaders,” said IBM Chairman and Chief Executive Arvind Krishna. “This is a pivotal moment in defining the future of AI.”

The AI Alliance consists of a broad spectrum of organizations working on various aspects of AI development, research, education, deployment and governance. It includes the creators of some of the most advanced open AI models in use today, including Llama 2, Stable Diffusion, StarCoder, Bloom and others.

Other members include the creators of AI frameworks, benchmarking and application development tools such as MLPerf, Hugging Face, LangChain and LlamaIndex. It also includes a host of universities and academic organizations at the forefront of AI research, as well as the builders of the hardware and infrastructure necessary to power AI development.

Red Hat said its upstream communities — including its Open Data Hub hybrid cloud development platform — will serve as an integration platform for models coming through the alliance. The company said it will also make this OpenShift and OpenShift AI platforms available for use by alliance members.

“Pursuing open innovation levels the playing field, allowing everyone to share in the benefits of generative AI,” said Jennifer Chayes, dean of the University of California at Berkeley’s College of Computing, Data Science and Society.

Andy Thurai, vice president and principal analyst of Constellation Research Inc., also highlighted the big names missing from the AI Alliance and added that many of the initiatives it’s pursuing are already being catered to elsewhere. For instance, tools like GAIA and MLPerf already enable developers to track the performance, accuracy and drift of their AI models. Meanwhile, offerings such as Fairlearn, AI Fairness 360 and the Accenture Fairness Tool already cover model fairness, he said, while popular benchmarking tools include DAWNBench, GLUE and SQuAD.

“I’m not sure what the Alliance is offering at this point, but there are some promises that have been made which are yet to be seen,” Thurai said. IBM might benefit a lot if other members of the consortium start using its AI chips, cloud and software for their own use cases, but I am very skeptical about the value add for other companies in this list.”

All told, the AI Alliance looks like it has good intentions, Thurai said, but it will only be able to make its impact felt if more of the industry’s major players join the party. “I would like to see the top ten players join this initiative to drive it forward, or at least see it join forces with others in some way,” the analyst added. “At this point, it comes across as more of an attention-seeking headline than anything concrete.”

The AI Alliance said the first item on its agenda is to form a number of member-driven working groups focused on achieving the goals outlined above. In addition, it will also create a governing board and technical oversight committee that will help to steer those working groups towards their goals while establishing overall standards and guidelines for each of its projects.

In addition to fostering collaboration among its members, the alliance will also work with governments, nonprofits and civil society organizations on various AI initiatives that are aligned with its goals.

“We believe it’s better when AI is developed openly – more people can access the benefits, build innovative products and work on safety,” said Meta President of Global Affairs Nick Clegg. “The AI Alliance brings together researchers, developers and companies to share tools and knowledge that can help us all make progress whether models are shared openly or not.”

Here’s the full analysis from Vellante and Strechay:

Image: Freepik