In a new essay, Meta CEO Mark Zuckerberg has called for an open-source future for generative AI. He argues that an open-source ecosystem benefits consumers around the world and fosters innovation, indirectly issuing a rhetorical challenge to closed-source developers (and Meta competitors) like OpenAI and Google.
“Meta’s business model is about building the best experiences and services for people,” Zuckerberg wrote. “To do that, we need to make sure we always have access to the best technology and that we don’t get locked into a competitor’s closed ecosystem that might restrict what we build.”
The principle behind open source models is that they allow for a collaborative approach to development. Unlike closed source models, which are limited to a company's internal team and its customers, open source models can be accessed, modified, and improved by anyone. For smartphone owners, the difference can be seen when comparing Apple's proprietary iOS operating system and Google's open source Android operating system. Android is therefore available on all types of devices, while iOS is limited to Apple devices.
In the field of AI, closed-source models, such as those from OpenAI and Google, are typically available only to a company’s customers and offer limited opportunities for modification. These models are designed to protect intellectual property and maintain control over the use of the technology.
According to Zuckerberg, developers benefit from having an open-source framework for AI. He argued that by making AI models available to a wide range of developers and researchers, the technology can be refined more quickly and effectively.
This would lead to improvements in both security and overall functionality, as more eyes on the code would help identify and fix flaws more quickly than in a closed system.
AI Open vs. OpenAI
A common counterargument against opening up AI is that it could be used maliciously by malicious actors or that adversarial nations could undermine the national security of rival countries with such technology. Zuckerberg acknowledged that, but countered by saying that closed-source models are not immune to those risks, as their code can also be stolen and used. Instead, he advocated for U.S. companies to work with the government and take an open-source approach.
Meta is already pursuing an open-source strategy with its Llama AI model portfolio. At the rehearsal, Zuckerberg announced the latest version, Llama 3.1, claiming that it is one of the most advanced models available thanks in part to its open source nature. Meta and Zuckerberg claim that by sharing such tools, they can better democratize AI development and ensure that its benefits are widely distributed. Zuckerberg believes that by making AI models open-source, a larger and more diverse group of developers can contribute to their development, potentially leading to more robust and secure systems.
The debate over open-source and closed-source models is likely to define the future of AI development. For now, it seems that they could coexist like other types of software, but the exact boundaries of that coexistence are far from defined.
Zuckerberg’s advocacy for open-source AI could influence other companies to adopt similar approaches and foster a more collaborative and inclusive AI development environment. However, the ultimate trajectory of AI development will depend on several factors, including technological advancements, regulatory frameworks, and market dynamics.
“I believe open source is necessary for a positive future for AI. AI has more potential than any other modern technology to increase people’s productivity, creativity, and quality of life, and to accelerate economic growth while driving progress in medical and scientific research,” Zuckerberg wrote. “Open source will ensure that more people around the world have access to the benefits and opportunities of AI, that power is not concentrated in the hands of a small number of companies, and that the technology can be deployed more uniformly and safely across society.”
