Unfortunately, it cannot learn a meta-language that has independence from context.
Deconstructing Deep Meta Learning
Carlos E. Perez

Or can it? That google model that combined many different application areas still learned a single representation. Could you combine a crap ton of models and learn an easy to learn intermediate representation, and from there train another model that takes this representation and learns any human-usable representation from there?

Like what you read? Give Laurens Hagendoorn a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.