metamorworks - stock.adobe.com
IBM and social media tech giant Meta have formed an AI alliance.
On Tuesday the two tech companies launched the new AI Alliance in collaboration with more than 50 founding members from other tech and educational institutions, including Anyscale, AMD, Cerebras, Cornell University, Dartmouth College, Dell, Stability AI and Oracle.
Notably missing from the list of alliance members are OpenAI, Microsoft, Google and AWS.
Still, the AI Alliance aims to create an open community that will accelerate responsible innovation in AI technology, while ensuring scientific rigor, trust, safety, security and economic competitiveness, according to IBM.
An alliance for openness
The AI Alliance launch arrived on the heels of a year that's seen growth in AI innovation with the mass consumerization of generative AI and a growing interest in open source tools such as Meta's Llama 2.
"The history of AI is full of the importance of open innovation, open science, research and open source technologies," IBM's Director of AI Open Innovation Anthony Annunziata said during an interview. "We believe that in order for AI to benefit society or for it to be safe and trusted, it needs to be open. We brought people to support that to build resources, to bring focus, to build, openly."
To ensure the AI Alliance accomplishes its goal of openness, its first focus is on safety and trust, Annunziata said. The safety and trust alliance working group has already started to meet.
"We're going to meet regularly, [and] we're going to identify priorities and existing tools," Annunziata said.
He added that other working groups focused on creating foundational tools and technologies will be created soon. They'll include a public policy and advocacy working group as well as a steering committee to operate the whole program.
Taking a working group approach is a great way to ensure the AI Alliance is effective, Futurum Group analyst Mark Beccue said.
"Working groups means you will get something done. Otherwise you're just kind of staring at each other," he said.
Pros and cons of open source
While promoting open source is good, there are pros and cons, Beccue continued.
Mark BeccueAnalyst, Futurum Group
"When open source is a viable option, it creates competition," he said. However, in tech, open source also has its downsides.
One weakness of open source is that it's not always secure. Enterprises that need tightly secured technology might choose a different route. Also, while technologies created within the open source community don't have the enterprise software price tag, they also don't have the support.
"It makes the assumption that a lot of enterprises are going to have experts that can look through all of the open source stuff and be able to use it correctly," Beccue said.
It's likely that many enterprises will use a combination of open source and closed source tools, instead of one or the other, he added.
Meta and generative AI
While the AI Alliance could encourage oversight of open source AI tools, it could also allow Meta to display how it is different from leading generative AI companies such as Microsoft, Google and AWS, said Chirag Shah, professor at the University of Washington Information School.
"The cynical way to look at this is that Meta missed this opportunity to be at the forefront of AI," Shah said. "So, they're kind of getting others to sign up with them collectively."
Microsoft, AWS and Google are associated with different alliances or groups that focus on responsible deployment of AI. For example, Microsoft partners with OpenAI, a research vendor that started out with a mission to deploy open source AI tools responsibly.
Google is part of the Partnership on AI, which looks to address questions concerning the future of AI. Microsoft, Meta and IBM are also a part of that group.
"I don't see what benefit [Google, Microsoft or AWS] will derive by signing up for AI Alliance," Shah said. "It's yet another thing, and it's heavily run by Meta."
Moreover, since the AI market will likely be dominated by one or two companies that enterprises already use, Meta and IBM could draw on the AI Alliance to remain relevant in the market, he said.
"It would be great if we do have this kind of alliance which focuses on doing things that are responsible," Shah added.
Digital workflow vendor ServiceNow is one of the vendors that joined the alliance. Part of the reason for joining is the vendor's continuous work of listening to customers and giving them the choice of large language models that meet their needs, according to Jeremy Barnes, vice president of product AI at ServiceNow.
"It's all about choice and transparency," Barnes said. "GenAI is a transformational technology that will have a profound effect on what enterprises are going to do in the future. ... Our participation shows that we're not going to go it alone with addressing these important aspects of generative AI."
Microsoft and Google did not immediately respond to request for comment.
Esther Ajao is a TechTarget Editorial news writer covering artificial intelligence software and systems.