An unequal distribution of compute power in academia is exacerbating inequality in deep learning, according to researchers at Virginia Polytechnic Institute and State University (Virginia Tech) and Western University.
The researchers reviewed 171,294 papers accepted for publication at large artificial intelligence (AI) conferences, in such categories as computer vision, data mining, machine learning, and natural language processing.
They noted that "AI is increasingly being shaped by a few actors, and these actors are mostly affiliated with either large technology firms or elite universities."
The researchers found that since the rise of deep learning, universities ranked 301-500 by U.S. News and World Report have published an average of six fewer papers at AI research conferences.
The findings, they said, offer "concrete evidence" of the need for a national AI research cloud.
Further, they contend the U.S. government should extend shared public datasets and other resources to help universities.
From Venture Beat
View Full Article
Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA