Over 50 entities, including universities, social media giants, startups and government agencies, co-operate and pool resources to form the AI Alliance. It’s the first interdisciplinary and intersectoral attempt at standardizing the AI environment of such magnitude. Regulation is still in its infancy e.g. here and here, and of course we don’t forget the…. regulatory superpower that is the EU.
On Tuesday, December 5th, Meta Platforms and International Business Machines (IBM) jointly launched the AI Alliance. The Alliance is comprised by over 50 AI companies and research institutions. This coalition, including members like Intel, Oracle, and Cornell University, advocates for an “open model” of AI. The alliance seeks to counter the dominance of closed, proprietary AI systems championed by companies like OpenAI.
MONTHS IN THE MAKING
IBM and Meta began working on this initiative in August, aiming to include organizations often overlooked in AI discussions. Furthermore, the AI Alliance members are keen to compete with the attention garnered by OpenAI and Microsoft. Since enterprises worldwide are spending nearly $16 billion on generative AI solutions in 2023, the alliance aims to provide an alternative to the closed systems prevalent in the market.
A USEFUL AND NECESSARY INITIATIVE
The launch aligns with the growing trend of businesses seeking multiple AI vendors to mitigate risks associated with a single provider. The alliance’s approach is more distributed and resilient, reducing dependency on specific institutions. Consequently, the focus of the AI Alliance includes areas like regulation, safety, and the imminent release of a benchmarking tool for AI safety and model validation. The alliance’s success will depend on its execution, particularly in offering integrated AI hardware, software, and tools to facilitate the use of multiple AI systems.
If you want to read the full manifesto as it has been published in IBM’ s site, including the full list of the founding members and their six common objectives (AI quality benchmarks, social responsibility, appropriate hardware development, AI skills building, education of society and openness) please click here.
