An algorithm that could 'predict' bits in arbitrary sequences which it had never been exposed to can hardly be called a 'learning algorithm'. You are confusing a class of 'innate' algorithms that could be used as a utilities for a learning program. If the infeasible idealization was feasible one would suppose that it would be a very powerful tool. There are some similarities between those kinds of algorithms (if any of my memories of your previous discussions are intact) and various Bayesian algorithms (which could be used in a range of learning situations) that can produce some sophisticated result that could model the learning that took place. for instance if the algorithm produced a number of interrelated Bayesian Networks that modeled what was learned that would be called a 'general learning algorithm'. (Not only Bayesian Networks of course, but that is an example.) Jim Bromer
On Sun, Nov 23, 2014 at 9:35 PM, Matt Mahoney via AGI <[email protected]> wrote: > On Sun, Nov 23, 2014 at 8:25 PM, Alan Grimes <[email protected]> wrote: >> An AGI is a **LEARNING SYSTEM**. Learning is it's sine qua non. If it >> can't learn, then it isn't an AGI, once you are at that point, all you >> need to do is connect a few network cables and walk away. > > I don't understand why you haven't created an AGI yet, if it's so simple. > > Oh wait, it's not. Are you familiar with Legg's proof that there is no > such thing as a simple, general purpose learning algorithm? Powerful > learners are necessarily complex. http://arxiv.org/abs/cs/0606070 > > In case you don't want to read the paper, here is a simple proof. > Suppose you have a small program with Kolmogorov complexity n that > learns to predict bits in arbitrary sequences. Then I can create a > sequence with Kolmogorov complexity slightly greater than n that your > program can't predict. My program simulates your program and outputs > the opposite of whatever your program predicts. > > -- > -- Matt Mahoney, [email protected] > > > ------------------------------------------- > AGI > Archives: https://www.listbox.com/member/archive/303/=now > RSS Feed: https://www.listbox.com/member/archive/rss/303/24379807-653794b5 > Modify Your Subscription: https://www.listbox.com/member/?&; > Powered by Listbox: http://www.listbox.com ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657 Powered by Listbox: http://www.listbox.com
