META And IBM Announce AI Alliance

By Brett Hurll

  In a groundbreaking move, Meta and IBM have spearheaded the formation of an AI Alliance, a conglomerate comprising industry giants like Intel and Oracle, alongside academic institutions such as Cornell University and the National Science Foundation. This alliance champions the cause of "open innovation and open science" in the field of Artificial Intelligence (AI). Rooted in the ethos of open source, where technology is freely shared, this coalition draws from a rich history of collaboration among major technology companies, academia, and a dedicated community of independent programmers.

Darío Gil, IBM's senior vice president and director of IBM Research, noted that since August, IBM and Meta have been instrumental in unifying entities not previously spotlighted in the AI discourse, unlike the widely known OpenAI. "The current dialogue on AI hasn't quite captured the diverse ecosystem enabling this AI era," Gil remarked, pointing out a gap this alliance aims to bridge.

The AI narrative has been predominantly shaped by generative AI since OpenAI introduced ChatGPT a year ago. Companies like OpenAI, Anthropic, and Cohere have been at the forefront, developing advanced, closed AI systems that are proprietary and require payment for use. The AI Alliance, however, consists of firms with their own AI products, yet they find themselves vying for attention in a market heavily influenced by OpenAI and its partner, Microsoft.

Enterprises globally are forecasted to invest approximately $16 billion in generative AI solutions this year alone, with the figure expected to soar to $143 billion by 2027, as per International Data Corp. This growth in generative AI is projected to far outpace overall IT spending in the coming years.

Despite a setback with its Watson system, IBM is making strides with the new Watsonx platform. Similarly, Meta, having fallen behind in AI advancements in recent years, is now focusing on its open-source AI system, Llama 2 AI model. The recent shifts at OpenAI have prompted businesses to seek alternative AI providers, reducing dependency on a single vendor.

"The AI Alliance’s launch is timely," Gil stated, emphasizing the resilience of a distributed approach to AI development, where no single institution can impede the progress of the collective endeavour.

Ritu Jyoti, group vice president of worldwide AI at IDC, believes this proposition could appeal to businesses looking for diverse AI vendors. Success, she notes, hinges on the Alliance’s execution, particularly its ability to offer an integrated suite of AI hardware, software, and user-friendly tools.

Advanced Micro Devices (AMD), aiming to challenge Nvidia in the AI chip sector, is committed to supporting an open AI ecosystem. Forrest Norrod, AMD’s executive vice president and general manager of its data center group, revealed plans to develop software enabling businesses to utilize its chips. AMD is poised to introduce AI accelerator chips, presenting a robust alternative to Nvidia’s products.

ServiceNow, another enterprise software firm with AI aspirations, is contributing to the AI Alliance. Its 50-strong AI research team will focus on the open scientific advancement of AI systems, underscoring the availability of choices for customers, as highlighted by Jeremy Barnes, ServiceNow’s vice president of product AI.

Gil from IBM concluded, “The future of AI won't be dictated by a handful of institutions. The open innovation world is one worth investing in.” The AI Alliance is concentrating on six key areas, including regulation and safety, and is set to unveil a benchmarking tool for AI safety and model validation soon.