RE: Supermathematics and Artificial General Intelligence
September 5, 2017 at 5:37 am
(This post was last modified: September 5, 2017 at 6:02 am by ThoughtCurvature.)
(September 5, 2017 at 2:28 am)Mathilda Wrote: Machine Learning is very popular right now because there is a lot of money to be made from statistical analysis of big data sets. This is what is driving the research, it is not the route to strong generalised artificial intelligence. But a new EU regulation (GDPR) comes into force in May 2018 that requires all automated decisions can be questioned and require an explanation. It's always been hard enough to convince clients to use a simple back-propagation three layer network that performs one statistical function, to explain how a deep learning model functions by extracting out hundreds of salient variables will be near impossible.
Like all AI techniques, machine learning has its limitations.
Mathilda, in contrast, research is going in directions largely concerning very general algorithms, or general intelligence.
Don't forget about unsupervised learning models that already exist today (and are only improving):
(1) Manifold learning or Deepmind's "Early Visual Concept Learning with Unsupervised Deep Learning"
(2) Generative Adversarial Networks that uses unsupervised learning (See Wikipedia "Generative_adversarial_networks")
etc.
EDIT:
What did you mean by three layers?
Don't forget about residual neural networks, and other stochastic models, that can do thousands of layers. (See 2016 arXiv paper: "Deep Networks with Stochastic Depth")
Even I myself, have configured optimally converging residual neural nets with 20 layers, with a lowed nvidia card.