Discussions surrounding the dominant voices in the AI zeitgeist typically do not include computing veterans such as IBM, Intel, Sony Group or Dell.
But on Tuesday, the four companies—along with young Meta, numerous top universities, and a host of technology startups and foundations—announced the establishment of the “Artificial Intelligence Alliance,” an apparent attempt to challenge the dominance of OpenAI and Microsoft. , Google, and most recently Amazon.
Darío Gil, senior vice president at IBM and head of the company’s research labs, said: “To some extent, but unfortunately for the most part, the discussions and conversations around artificial intelligence last year were dominated by Concentrated on a very small number of institutions.”, told wealth. “The reality is, the field is much bigger than that.”
When asked who the “very few institutions” were referring to, Gill declined to specify: “You know who.”
“Open” artificial intelligence
The formation of the AI Alliance continues a long-running debate among developers over the values of “open” and “closed” development of artificial intelligence.
Despite its name, OpenAI, the creator of ChatGPT, has kept its model or vast artificial intelligence algorithm secret. Developers can access them only with a license from OpenAI, which counts Microsoft as its biggest backer. Google, another AI front-runner, and Amazon, which recently unveiled its ChatGPT solution and invested in popular AI startup Anthropic, also don’t open source or let researchers fully download their models. All tech giants cite competition and security as reasons for why they lock down technology.
This austerity policy has caused consternation in the research community and competing companies. (Indeed, competitors watched with glee as OpenAI’s corporate leadership descended into disarray in November.) “There’s a lot of debate about: Should the future of AI be closed and proprietary? Or should it be open source, open science, and What is the role of open innovation in this area?” said Gill, an IBM executive.
The Artificial Intelligence Alliance falls into the latter camp. The group of more than 50 people is united around a number of broad goals, including creating a common framework for assessing the strength of AI algorithms, investing in AI research grants, and collaborating on open source models.
In addition to corporate giants, other players include chipmakers AMD and Cerebras, artificial intelligence startups such as Hugging Face and Stability AI, and Ivy League schools such as Yale, Cornell and Dartmouth.
Gil cited, for example, a recent collaboration between IBM and NASA to develop an open-source artificial intelligence model trained on geospatial data, which he said could help track deforestation or predict crop yields. He also said that IBM has committed about $100 million to universities to support artificial intelligence research projects over the next five years, and that the computing giant has partnered with Meta to build an open source toolkit for artificial intelligence development.
As for governance, Gill said the league is still working out the details. So far, the focus has been on building a coalition and setting the goals of the organization. Next steps include forming a “technical working group” of more than 50 participants and designing a governance structure that could lead to an external nonprofit.