wait why does it hold back machine learning? I feel that recent transcription models, genome approximators and classifiers are quite something and large-scale experiments on new problem spaces are green-lit much quicker than before. i personally feel that because the space is so big, there are many more attempts made to use ML systems using those same modern architectures to help in new domains or optimize existing ones.

I am very irritated by the focus on language models and image / video /audio predictors and feel that too many resources are being used exclusively for those. The overall impact is obviously trampling down any of the good stuff.