Hugging Face CEO Clément Delangue doesn’t like the tendency of big tech companies to hold on to machine learning codes like business patents. Instead, he’s working to build a world where every developer or startup can freely access artificial intelligence models—and helping AI, effectively, go open-source.
Today Hugging Face is democratizing the development of the most consequential technology of our age. The platform serves upwards of 50,000 organizations using its services, including more than 300,000 models, 100,000 applications, and 50,000 datasets. And its approach stands in sharp contrast to other AI developers, who heavily guard their proprietary technology and models.
“We need more companies and organizations to share their models and datasets publicly and in open-source so that everyone can understand and build AI themselves,” Delangue tells Quartz.
Delangue co-founded Hugging Face in 2016, with the company initially imagined as a chatbot for teenagers before upgrading to being an open-source provider of Natural Language Processing (NLP) technologies. The platform would later start open-sourcing popular large language models. “I think through the open source model, you can do things a bit differently … you can, as a startup, empower the community in a way, and create a thousand times more value than you would by building a proprietary tool,” Delangue said in a 2021 interview.
This fall, a team of Stanford researchers issued a report measuring tracks how much information AI makers like OpenAI, Meta, and Google disclose about the building of their models—and declared that none of the most prominent developers are sharing enough information about their potential impact on society. Delangue says this year Hugging Face strove to make that issue its calling card. “One big achievement was also advocacy for more openness and transparency with policy-makers,” Delangue says.
Delangue credits transparency as a reason Hugging Face became a widely-used open platform for AI builders—and its approach has struck a note beyond developers, too. Like his fellow tech executives, Delangue has testified about AI on legislative floors. Speaking at a US Congress hearing in June, he underlined how the AI field is dominated by a few rich companies “who give limited or no open access to novel AI systems, including those based on open research.”
But even if Hugging Face critiques the big tech behemoths shrouding their AI tools in secrecy, it’s found a way to work with them, too. Hugging Face’s pitch has attracted heavyweight investors including Google, Microsoft, Amazon, IBM, Nvidia, and Qualcomm. In August, it was valued at $4.5 billion.
Hugging Face sees itself as the community building the future of AI development, its platform having already drawn comparisons to a Github for AI. In 2024, Delangue is eyeing a growth milestone: 10 million users on the platform. As AI races onward, Hugging Face is shifting developers towards something resembling a community effort—powering the way to system-sharing, collaboration, and feedback.
This story is part of Quartz’s Innovators List 2023, a series that spotlights the people deploying bold technologies and reimagining the way we do business for good across the globe. Find the full list here.