AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Is There a Dycodomy Between Continuous Indiscreet and Continuous InDiscrete?
There is this apparent dycodomy between continuous indiscreet and discrete computations. It's almost impossible to train any neural network of fixed depth to do the same thing. And we're always going to have hybrid systems that kind of comb combine both types of reasoning. Is it possible to project that discreet computation fully into a analogue type space, if you just have enough, you know, perimeters or something?