ML model evaluation techniques may be hiding new developments in AI (arielos.substack.com)

Choosing the wrong evaluation method may lead us to subobtimal models, while there may be new architectures (the author suggests to look back to neuroscience for inspiration) that can better handle what we expect from future models