What about Cyc? It operates on a taxonomy (graph) which is a highly complex topology. ~PM
Date: Mon, 22 Jun 2015 18:57:05 -0700 From: [email protected] To: [email protected] Subject: Re: [agi] Re: Could Brain Emulation be NP-Hard? I seldom ask this as I usually work at the level of abstractions. But could you please give some examples of topological relationships that are difficult to express computationally? I am not sure I follow exactly what you have in mind. Thanks. - samantha On 06/16/2015 06:09 PM, J. Andrew Rogers wrote: On Jun 16, 2015, at 3:26 PM, Dean Pomerleau <[email protected]> wrote: In short, growing evidence supporting the importance of cortical oscillations in neural processing suggests that this sort of analog/digital feedback loop might be critical to how the brain works, and that such interactions might be very hard (possible intractably hard) to model accurately (i.e. emulation vs. merely crude simulation) on a digital computer, in a similar way to how protein folding is intractable to model on a digital computer. The tractability challenges of computational dynamics for brain-like models is related to why we can’t analyze the dynamics of *any* non-trivial physical world system. It is not coincidence that all “big data” computation focuses solely on relationships in the electronic world and not the physical world. Interestingly, computer scientists rarely notice that these software systems do not exist until you point it out. And when you do point it out they are at a loss to explain why. It is only “obvious” in hindsight. Virtually all existing computer science is based on the manipulation of graph-like data models and primitives. The problem is that some systems, notably physical world systems, have relationships that are fundamentally topological in nature. Graphs are a special, strict subset of more general topological computing representations; it is not possible to construct a scalable topological computational model on top of graph primitives. There is no computer science literature for computing on topological data models. To the extent algorithms and data structures exist to handle basic topological data models (e.g. R-trees), they exhibit pathological scalability because they are shoehorned into traditional graph models. If you want to compute on topological models at scale, you need to build a completely new computer science stack, from the most elementary primitives on up. And it needs to have an efficient implementation on conventional silicon. If you can directly manipulate topologies as computational constructs, instead of graphs only, many types of computational dynamic suddenly become *much* more tractable. In practice, the use of inappropriate algorithms and data structures to represent topological relationships are responsible for most intractability related to expressions of physical world system dynamics on a computer. It just never crosses the mind of most computer scientists working on such things and it is never discussed in computer science curricula. AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription ------------------------------------------- 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
