The open-source AI debate is heating up in Big Tech, with recent headlines from Google and Meta.
According to a report by CNBC, Google’s newest large language model (LLM) PaLM 2 uses almost five times more text data for training than its predecessor. However, Google has been unwilling to publish the size or other details of its training data, causing frustration among Google engineers. Dmitry (Dima) Lepikhin, a senior staff software engineer at Google DeepMind, even tweeted: “whoever leaked PaLM2 details to cnbc, sincerely fuck you!”
Meanwhile, Meta’s chief AI scientist Yann LeCun did an interview with the New York Times, emphasizing Meta’s open-source AI efforts. LeCun believes that the platform that will win will be the open one, and that the growing secrecy at Google and OpenAI is a “huge mistake” and a “really bad take on what is happening.”
While Meta has already given away one of the most critical AI/ML tools ever created, PyTorch, the company still has a nuanced take on open source. Joelle Pineau, VP of AI research at Meta, believes that accountability and transparency in AI models is essential, but some levels of openness go too far.
Despite the debate, LeCun remains committed to open-source AI, saying “progress is faster when it is open.”
In recent years, Artificial Intelligence (AI) technologies have revolutionized the way we interact with the world. This is due in part to the emergence of open-source AI initiatives, which aim to make AI more accessible and cost-effective. One particular open-source AI initiative that has gained traction in recent years is Google’s Meta project.
Google’s Meta project began in 2018 as an effort to make AI technology more available to mainstream audiences. It provides developers with libraries of code and learning models that are open source and free to use. By doing so, Meta hopes to make AI more accessible and easier to implement in a wide range of applications.
In addition to Google’s Meta project, Google News has played a important role in driving the open-source AI movement. Google News collects and organizes breaking news from around the world, making it easier for developers and entrepreneurs to stay up-to-date on the latest AI advancements. Google News also provides developers with tutorials and resources to help them build their own AI-based applications.
The combination of Google’s Meta project and Google News has been instrumental in fueling the open-source AI movement. By making AI technologies more available and accessible, Meta and Google News have enabled more businesses and individuals to get involved with AI development and promote the benefits of open-source AI technology.
It is clear that Google’s Meta project and Google News have been critical to driving the open-source AI movement over the past few years. By providing access to free learning models and up-to-date information, Meta and Google News have enabled more developers and entrepreneurs to implement AI in their projects and products. As a result, the open-source AI movement has grown and continues to expand and influence the tech world.
