Alphabet Inc’s cloud computing arm (GOOGL.O) has announced a partnership with startup Hugging Face, a move strategically geared to streamline Artificial Intelligence (AI) software development within Google Cloud,
The collaboration comes in response to the increasing interest of businesses, both large and small, to deploy AI for specific purposes such as enhancing existing products or fulfilling internal needs.
The partnership addresses the evolving landscape of AI, as companies seek to create software optimized for specific tasks and requirements.
Hugging Face’s Chief Executive, Clem Delangue, noted in an interview with Reuters, “It shows we are moving into a world where there’s not just one model, but there are as many models as companies. Every company is training, fine-tuning and optimizing their own models, which is the direction that the field is taking.”
With the recent launch of generative AI tools by major players like Microsoft and Google’s parent company, Alphabet, there is a growing demand for AI computing capabilities in the cloud.
Google Cloud CEO Thomas Kurian mentioned that the trajectory of cloud-based AI computing suggests it may surpass the traditional cloud software market in the future.
The partnership between Google Cloud and Hugging Face involves a revenue-sharing agreement, although specific terms were not disclosed. This collaboration enables developers using Google Cloud’s technical infrastructure to access and deploy Hugging Face’s repository of open-source AI software, including models and data sets.
Hugging Face, based in New York, has become a central hub for open-source AI software and previously entered a similar arrangement with Amazon.com. Spend less. Smile more. ‘s cloud unit in February last year.
The partnership seeks to meet this demand as cloud customers increasingly express interest in modifying or building their own AI models rather than relying on off-the-shelf options,
This includes considerations such as implementing specific safety frameworks to secure against data leakage or evaluating the quality of the running model.
Reflecting the surging interest in building AI software, Hugging Face anticipates doubling the size of its repository of AI-related software within the next four months—a remarkable acceleration compared to the four years it took to reach its current state.
Developers utilizing Google Cloud will have access to an advanced version of Google’s in-house-built AI chips, or tensor processing units. Furthermore, Google plans to complete the technical integration to use Nvidia’s coveted advanced AI chips, the H100, within a matter of “weeks,” according to Kurian.