Stanford University’s Institute for Human-Centered AI inked a memorandum of understanding with Switzerland’s newly formed National AI Institute—jointly led by ETH Zurich and EPFL—on the sidelines of the World Economic Forum in Davos. The pact targets open-source, open-data foundation models and shared evaluation benchmarks, aiming to bolster transparency, reliability, and access while reinforcing academia’s role in shaping the AI agenda. Leaders from the institutions framed the tie-up as a counterweight to big tech’s influence, emphasizing models that advance science and reflect societal values and cultural diversity. The collaboration envisions joint research, educational exchanges, workshops, and shared compute, positioning the transatlantic alliance as a hub for rigor and responsibility in the fast-evolving field.
Related articles:
BLOOM: A 176B-Parameter Open-Access Multilingual Language Model
Introducing Llama 2
LAION-5B: An Open Large-Scale Dataset for Training Foundation Models
Vicuna: An Open-Source Chatbot Impressing GPT-4
