Open Source Software (OSS) is a proven model that delivers tangible benefits to businesses, including improved time-to-market, reduced costs, and increased flexibility. OSS is pervasive in the technology landscape and beyond it, with adoption across multiple industries. In a 2022 survey by Red Hat, 95 percent of IT leaders said they are using open source in their IT infrastructure, which will only increase.
Artificial intelligence (AI) is no different from any other technology domain where OSS dominates. In a recent paper published by Linux Foundation Research, written by Dr. Ibrahim Haddad, General Manager of the LF AI & Data Foundation, over 300 critical open source projects have been identified offering over 500 million lines of code, contributed by more than 35,000 developers who work side by side to advance the state of technology in an open, collaborative, and transparent way.
As with other industries, OSS adoption in the AI field has increased the use of open source in products and services, contributions to existing projects, the creation of projects fostering collaboration, and the development of new technologies due to this amazing success story.
In this paper, you will read that while AI in open source has followed a similar model to other industries embracing the popular methodology, Dr. Haddad has some unique observations to share, which include:
An incubation model for AI open source projects is effective when appropriately executed by neutral organizations that can scale them, such as the Linux Foundation.Consolidation is bound to happen around platforms, frameworks, and libraries that address similar challenges. Unlike typical fragmentation scenarios, where there are winning and losing projects, Dr. Haddad believes the net result will be a win-win as successful projects grab their share of contributors.License choices can affect a project’s growth — and licenses approved by the Open Source Initiative (OSI) are most preferred because developers and enterprises are already familiar with them.Open data licenses such as Community Data License Agreement (CDLA) have begun to commoditize training data. These license terms will help democratize the overall AI marketplace by lowering the barriers to entry when offering an AI-backed service. Proprietary datasets will continue to exist, but data availability under the CDLA licenses (two versions exist) should allow everyone to build credible products, including smaller players.
So what does this mean for the future of AI? It means that businesses will continue to rely on open source software to power their AI initiatives and that collaboration will be key to success. The open source model has been successful in AI because it allows developers to come together and share code, data, and ideas. This type of collaboration is essential for advancing any technology, and we can expect to see even more impressive innovations come out of the AI community in the years to come. Ultimately, we are faster and more innovative together.