AIGricea's pitch for more reproducible AI research
A post promoting Gricea claims 97% of recent conference papers had replication issues, presenting the service as a way to address the problem.
TLDR
A post argues that building cumulative knowledge about AI's impact on people requires reproducible studies and evaluations. It claims 97% of recent conference papers had replication issues, promotes Gricea as a solution and links to an arXiv paper.
Gricea's pitch for more reproducible AI research
A post promoting Gricea claims 97% of recent conference papers had replication issues, presenting the service as a way to address the problem.
