Meta, in conjunction with International Business Machines (IBM), has created an AI Alliance—a coalition of AI firms and research institutions pushing for a responsible and open approach to AI development.
The AI Alliance wants to foster an open community that will allow developers and researchers to speed up innovation in the sector in a responsible way while ensuring safety, security, diversity, and economic competitiveness.
The AI Alliance, which is a cooperation of 50 tech companies, founding members, and collaborators globally, aims to nurture a vibrant AI hardware accelerator “ecosystem by boosting contributions and adoption of essential enabling software technology.”
An open approach
While other top AI firms like OpenAI, Microsoft, and Google are into keeping models under tighter controls, IBM and Meta think an “open approach makes more sense.”
OpenAI and rivals like Anthropic have led the charge in advanced AI model development, but built as closed or proprietary systems managed by their makers. These require companies to pay for their use.
However, many of the Alliance’s members have their own AI products and are battling to catch up with the attention rush that OpenAI and its investment partner Microsoft are getting, according to the Wall Street Journal.
The firms in the Alliance therefore want to promote “open” alternatives to OpenAI, the maker of ChatGPT.
Both Meta and IBM have been active in AI, although IBM’s own history “has been marred by the failure of its Watson system.” Meta has also created its own models and attempted an open-source AI system through its Llama 2 AI model.
Chip maker: Advanced Micro Devices (AMD) executive vice president Forrest Norrod said they support an open AI ecosystem with their hardware and will build software that will allow businesses to use their chips.
AMD also wants to break Nvidia’s dominance with a “strong” alternative.
Responsible creation and use
According to Business Insider, the AI Alliance plans to develop and deploy benchmarks, evaluation standards, and resources that allow for responsible creation and use of AI systems at global scales. This includes the creation of a catalog of security and trust tools.
“This is a pivotal moment in defining the future of AI,” said IBM chairman and CEO Arvind Krishna.
“IBM is proud to partner with likeminded organizations through the AI Alliance to ensure this open ecosystem drives an innovative AI agenda underpinned by safety, accountability, and scientific rigor.”
This collaboration of AI tech stakeholders is also expected to unveil initiatives that inspire the open development of AI in safe and beneficial ways.
It should also host events to explore AI use cases and display how Alliance members are using open technology in AI responsibly and for good.
Among the AI Alliance members are AMD, Cornell University, CERN, Dell Technologies, EPFL, Imperial College London, ETH, Intel, NASA, Linux Foundation, NSF, Oracle, Red Hat, Partnership on AI, Sony Group, and Stability AI.
President of global affairs at Meta, Nick Clegg, said the AI Alliance brings together all these stakeholders,s like researchers and developers, to share tools as well as knowledge that can help the sector grow. According to him, models can be shared openly or not.
AMD CEO and chair Lisa Su emphasized transparency across all aspects of the rapidly developing AI systems.
“We can help ensure the transformational benefits of responsible AI are broadly available,” she said.
The AI Alliance is also looking at creating more model options in the AI sector. Following the upheaval at OpenAI late last month, businesses want to narrow the risk of relying on a single vendor.
“This other way, it’s a much more distributed approach, but much more resilient, because no given institution can derail the success of the open engine,” said Darío Gil, senior vice president at IBM and director of IBM Research.
As such, businesses want to explore other AI systems as viable alternatives.