Image: AI Alliance Symposium 2023.
The alliance for AI – AI Alliance in its original version – initiated by IBM and Meta Platforms (formerly Facebook), which brings together more than 50 companies and universities from around the world among its founders, was launched on December 5. Its goal ? Develop and promote open and reliable artificial intelligence.
After its creation, the alliance gathered on December 14 for the AI Alliance 2023 symposium, intended to promote its activities. Will this group, which aims to act as a counter-power to Microsoft/OpenAI and Google, which are currently leading the way in the field of generative AI, be able to establish itself as a third force?
During the gathering, several of its members spoke about their vision of current AI and its future developments.
For transparent and reliable AI
“AI is today at a major turning point. While it was previously a technology created by specialist developers and used only by people with a high level of training, it can now be created by anyone and used by anyone. This opens up new possibilities for social and societal change, as well as new ways of working,” recalls Akio Yamaguchi, CEO of IBM Japan.
“In this context, it is very important that AI is a transparent and trustworthy technology. We launched the AI Alliance in the hope of achieving such AI with as many people as possible. Together, we want to create AI that is useful to society and builds the future,” he says.
For AI regulation
“The year 2023 saw the development of international rules on AI. But, from 2024, it will be important to know how to firmly protect it and how to prevent its misuse and misinformation,” warns Shoji Watanabe.
Before adding that generative AI and its language models, which have been particularly highlighted this year, will be increasingly used in various industries and in many companies. For him, “issues of security and reliability will become extremely important, and it will be necessary to promote the creation of an open environment to achieve this”.
However, the debate is tenacious between supporters of open AI and those of proprietary AI. “I believe that these are not opposing concepts and that the two should be combined, or used on a case-by-case basis,” he maintains.
For open source AI
On this subject, Dario Gil, senior vice president of IBM and director of IBM Research, also spoke: “It is very important that we do not have to choose between open source and proprietary. If we look at the history of information technology, we see that there have been similarities in the past with what we see today. Take, for example, computer operating systems. The free software community gave birth to Linux, which has now established itself as a server operating system and enjoys the support of many users. »
Before explaining what this grouping means for him: “The AI Alliance includes world-renowned companies and universities, which will carry out co-creation activities in various projects in the future. So far, the social debate on AI has been insufficient. Now it is important to constructively advance these discussions and take action, and the AI Alliance will lead this movement. »
Source: ZDNetJapan