Meta’s Global Head of Business Nick Clegg Defends Open Source AI Approach

Meta (META)‘s approach publishing open source AI models It alleviates concerns about the potential excessive concentration of power in the hands of a few large tech companies. Nick Cleggglobal business president of the company. Unlike its competitors Open AI, Google (GOOGL) And AnthropicMeta’s path in the AI ​​arms race has largely been defined by its open source nature, which allows the source code for models like the recently released Llama 3.1 405B to be freely downloaded and modified. Unlike closed AI systems, Meta’s open source offerings “really democratize access to new technology quite significantly,” Clegg said while speaking at the Concordia Summit in New York today (September 24).

While Meta has been making its AI models available for commercial use for some time now, Llama 3.1 405B, released in July, is the largest open-source AI model released to date. According to the company, the new model’s capabilities are comparable or superior in long-context benchmarks when compared to competitors like OpenAI’s GPT4o and Anthropic’s Claude 3.5 Sonnet.

This isn’t the first time Clegg has defended Meta’s open-source approach. A former UK member of parliament who initially joined Meta in 2018 as a lobbyist and communications officer, much of Clegg’s current role as head of global affairs involves: Defending Meta’s productsIn a column published in the Financial Times last year, Clegg wrote, Reiterated belief in Meta being open source It will offer broader access to AI rather than keeping the technology concentrated among powerful companies.

Open source AI models should come as no surprise

Open source isn’t a particularly revolutionary concept in the tech world. “If you look at the evolution of the internet in the old days, or if you look at the encryption industry or the cybersecurity industry, they’re all built on open source technology,” Clegg said, noting that open access has long been part of Meta’s corporate strategy. “Over the last decade, our AI research team has open sourced over 1,000 AI databases and models. So, it’s nothing new to us,” he added.

Clegg said Meta’s decision to go open source also fits its structure. Because the company’s business model doesn’t rely on charging for access and instead relies on advertising, Meta can open source its products without negative financial consequences, the executive explained.

Despite the many benefits of open-source software, critics worry that the practice could create opportunities for exploitation by bad actors. Such fears are unfounded, according to Meta CEO Mark ZuckerbergArguing that it is better for technology to be open source so that big actors can check the power of smaller, bad actors.

Clegg argued that broader access to AI ultimately improves the technology itself. “You gain wisdom from people testing and retesting,” he said, rather than allowing companies to “play ‘Whack-a-Mole’ with their own models.” “Experience, particularly over the last 20 years, shows that an open approach generally leads to more innovation, more creativity, more democratization of technology, and more resilient systems over time.”