Illustration: Annelise Capossela/Axios
The government Monday announced the first round of projects that will access the National Artificial Intelligence Research Resource pilot.
Why it matters: The projects will aim to advance deepfake detection, AI safety and medical diagnosis — some of the most pressing issues policymakers point to for advancing the technology responsibly.
Driving the news: The National Science Foundation and Energy Department announced 35 projects that will be supported through the NAIRR pilot.
- They'll receive access to advanced computers to tackle critical AI priorities.
The big picture: NAIRR is meant to help expand access to AI so that not just tech giants can use powerful technology.
- In addition to having computational access, AI developers such as Hugging Face say knowledge sharing and collaboration are crucial for AI democratization.
- Other smaller open source actors pushing for AI openness in the NAIRR include EleutherAI, Databricks and AI2.
What they're saying: NSF director Sethuraman Panchanathan said the agency plans to rapidly expand partnerships and secure the investments needed to realize the NAIRR vision.
- NAIRR "is a fantastic endeavor that will reach its full success if it can bring more diverse perspectives and more openness to the study and development of AI systems," Yacine Jernite, Hugging Face machine learning and society lead, said in a statement.
What's next: The next application opportunity for NAIRR was opened today, including for AI research related to cloud computing platforms, access to foundation models and privacy-enhancing tools.
- This round includes resources from Microsoft, Amazon Web Services, NVIDIA, OpenAI, Anthropic and other private sector partners.
