Will the Meta-Ibm alliance change the development of artificial intelligence?
Meta and IBM have launched a coalition of more than 50 companies and research institutes who want a so-called open model for artificial intelligence. Facts, announcements, promises and problems (with Biden?)
IBM and Meta have created an alliance on artificial intelligence, starting from an open source approach.
The IT giant and the group that brings together the social networks Facebook, Instagram and WhatsApp have launched a coalition of over 50 companies active in the field of artificial intelligence and research institutes who want a so-called open model for artificial intelligence.
Despite being a slow process for decades, AI became popular almost overnight when just a year ago OpenAi launched ChatGPT .
Part of the confusion around open source AI is that, despite its name, OpenAI – the company behind ChatGPT and the Dall-E image generator – builds decidedly closed AI systems, the Guardian points out.
The AI Alliance said it is pooling resources to support “open innovation and open science” in artificial intelligence.
The founders therefore include Meta and IBM but also Intel and Oracle together with institutions such as Harvard and Imperial College London. All members of the AI Alliance largely support open source, an approach in which technology is shared for free and is based on a history of collaboration between Big Tech, academics, and a fervent movement of independent programmers.
That leaves Google, Amazon, Microsoft-backed OpenAi on one side. It's all about open source innovation (Linux as opposed to Windows to name one example).
“It's kind of a classic regulatory capture approach trying to raise concerns about open source innovation,” “I mean, this has been the Microsoft model for decades, right? They have always opposed open source programs that could compete with Windows or Office. They're taking a similar approach here,” said Chris Padilla, who leads IBM's global government affairs team.
The move by Meta and IBM could encounter difficulties given the approach of the White House, an analyst points out to Startmag : “When Biden issued the executive order on AI he underlined that making certain parameters public can indeed bring substantial advantages for innovation, but also substantial security risks."
All the details.
WHAT THE AI ALLIANCE PROMOTED BY META AND IBM WILL DO
The AI Alliance has described its mission as to accelerate and spread open innovation across the AI technology landscape, to improve core capabilities, safety, trust in AI and to responsibly maximize the benefits for people and society everywhere. This group brings together a critical mass of computational resources, data, tools and talent to accelerate open innovation in AI.
THE ADVANTAGES
According to this new AI alliance, shared work in this modality offers three main advantages. First of all, speed: sharing models will allow researchers to build on the work of others, accelerating progress in the field of AI. Second, security: allowing independent groups to review code created by others is the best way to identify potential flaws and risks. Finally, equality of opportunity: by providing everyone with access to developed tools, a level playing field is created where independent researchers and startups have the same opportunities as well-funded companies.
NASAM CERN AND IMPERIAL COLLAGE ALSO AMONG THE FOUNDERS
Founders include public and private organizations, researchers and academia: AMD, Anyscale, CERN, Cleveland Clinic, Cornell University, Dartmouth College, Dell Technologies, EPFL, ETH, Hugging Face, Imperial College London, Intel, INSAIT, Linux Foundation, MOC Alliance – managed by Boston University and Harvard University, NASA, NSF, Oracle, Red Hat, Roadzen, ServiceNow, Sony Group, Stability AI, University of California – Berkeley, University of Illinois, University of Notre Dame , University of Tokyo and Yale.
THE META ATTACK ON THE CLOSED MODEL
Supporters of open source AI technology, made public by developers for others to use, see the approach as a more efficient way to cultivate highly complex systems. In recent months, Meta has released open source versions of its large language models, which form the foundation of AI chatbots.
Meta's chief AI scientist, Yann LeCun, recently took aim at OpenAI, Google and startup Anthropic on social media for what he described as "massive corporate lobbying" to write the rules in a way that benefits consumers. their high-performance artificial intelligence models and could focus their power on developing the technology. The three companies, along with key OpenAI partner Microsoft, have formed their own industry group called the Frontier Model Forum.
“In a future where AI systems are poised to be the repository of all human knowledge and culture, we need platforms to be open source and freely available so everyone can contribute,” LeCun wrote. “Openness is the only way to make AI platforms reflect the totality of human knowledge and culture.”
THE FIRST COMMENTS
“AI has the potential to save the world, end the world and make many important leaders very rich” observes Finimize adding that “as concerned that some of the largest companies in the sector may be less focused on the former, some of the minds smarter have come together with the mission to develop AI solutions safely, slowly and openly.”
“With much of the world worried that the breakneck speeds of AI developers could end in a crash, this slow and steady force could attract quite a few fan clubs,” concludes Finimize.
This is a machine translation from Italian language of a post published on Start Magazine at the URL https://www.startmag.it/innovazione/meta-ibm/ on Fri, 08 Dec 2023 06:15:31 +0000.