In a surprising move, established tech giants IBM, Intel, Sony, and Dell, along with Meta, chip manufacturers AMD and Cerebras, and AI startups Hugging Face and Stability AI, have come together to announce the "AI Alliance" on Tuesday. This move is seen as a direct response to the growing influence of OpenAI, Microsoft, Google, and Amazon in the AI landscape.
IBM's CEO, Arvind Krishna, described it as a pivotal moment for AI, stressing the importance of safety and accountability. "IBM is proud to collaborate with like-minded organisations through the AI Alliance, ensuring an open ecosystem that drives innovation in AI with a focus on safety, accountability, and scientific rigour."
The alliance aims to prioritise responsible AI development, including the creation of safety and security tools, as mentioned in their official statement. They also plan to increase the availability of open-source AI models, moving away from closed, proprietary systems.
Why this alliance now
The AI Alliance's creation sparks an ongoing discussion among developers regarding the merits of "open" versus "closed" artificial intelligence development. Despite its name, OpenAI, the entity behind ChatGPT, has maintained tight control over its models and AI algorithms. Access to these resources is restricted, requiring explicit permission from OpenAI. Similarly, other AI leaders like Google, Anthropic, and Amazon, have not made their models open-source or fully downloadable for researchers.
The tech giants, including OpenAI, Google, and Amazon, claim competition and safety concerns as the driving factors behind the restricted access to their technology.
Several members of the alliance, companies with their own AI products, find it challenging to compete with the attention garnered by OpenAI and its investment partner. For many, the growth is fueled by revenues from enterprise technology. Research firm IDC predicts global spending on generative AI solutions to reach nearly $16 billion this year, with projections soaring to $143 billion by 2027.
Despite IBM's AI history being overshadowed by the Watson system's failure, the company is making a comeback with its latest Watsonx system. Meta has created its own AI models. The tech giant aims to assert its presence as an open-source AI system, utilising its Llama 2 AI model.
AMD, the chip maker aiming to challenge Nvidia’s dominance in AI chips, will back an open AI ecosystem with its hardware. Lisa Su, AMD CEO and Chair said, “The history of our industry highlights how open, standards-based development leveraging the capabilities of the entire industry both accelerate innovation and ensure technology advances have the largest positive impact.
Jeff Boudreau, Chief AI Officer at Dell Technologies, emphasises the necessity of open innovation and ecosystems for AI progress benefiting humanity.
Kevin Murphy, Chief Science Data Officer at NASA, underscores that open innovation is crucial for equitable access and collaboration in AI, rooted in principles of diversity, trust, and ingenuity.
ServiceNow, with its own AI aspirations, announced its 50-person AI research team's participation in the AI Alliance. CJ Desai, President & COO of ServiceNow, emphasised the importance of a diverse approach in developing responsible and secure Large Language Models (LLMs), rejecting a winner-take-all strategy.
How the AI Alliance will work
The AI Alliance will kickstart its efforts by creating working groups led by members in key subject areas. It will set up a governing board and a technical oversight committee focused on advancing project areas and establishing overall standards and guidelines.
Beyond uniting top developers, scientists, academics, students, and business leaders in the AI field, the AI Alliance aims to collaborate with significant existing initiatives from governments, non-profits, and civil society organisations engaged in valuable and aligned work in the AI domain.