> > And the power of microbial intelligence is based on inter-agent > communication...
Then looking at Cyc's complexity it's difficult to judge since the pure > data value is not as important as it's interrelational operability in human > interactions so a purely k-complexity view doesn't reflect informational > value perhaps.. That's the important point. It's also why we're so successful as a species. I would say communication is far more important than IQ. Even a mindless buffoon can do reasonably well in life, if s/he is able to take advice and follow directions decently. Every time we discover something new and useful, whether intentional or through serendipity, it becomes a part of our knowledge heritage, which means that even without the ability for intellectual reasoning, our use of memes still represents a *powerful* distributed learning algorithm. If we had a system that could effectively generate (or event just extend) an ontology through natural language interaction or processing, then it could plug into that distributed learning algorithm and take advantage of the knowledge base accumulated by it, the product of countless minds and experiences. Hand-coding of an ontology as is done with Cyc would be a waste of time, and of little value, because the knowledge itself is not where the value actually lies, but in the ability to acquire, accumulate, and share knowledge. On Sun, Feb 15, 2015 at 9:12 AM, John Rose via AGI <[email protected]> wrote: > > -----Original Message----- > > From: Matt Mahoney via AGI [mailto:[email protected]] > > > > On Thu, Feb 12, 2015 at 9:44 PM, John Rose via AGI <[email protected]> > > wrote: > > > > > > The k-complexity of a single bacterium’s molecular disassemblage has > > greater magnitude than that of OpenCyc’s ontological delineated totality. > > > > > > Yea or nay? > > > > OpenCyc 4.0 has 239,000 terms and 2,093,000 triples according to > > http://www.cyc.com/platform/opencyc > > ReserachCyc has 500,000 concepts and 5,000,000 assertions. Each > assertion, > > expressed as a triple encoding 3 concepts at log(500,000) = > > 19 bits each would be 300M bits. Assuming a Zipf or power law > distribution > > over the concepts would probably allow the database to be compressed to > > 150M bits. > > > > The genome of E. Coli is 4M base pairs or 8M bits assuming no compression > > (it compresses poorly). > > > > The genome is a starter for estimating an organisms complexity. But I > would suggest as a genome is "injected" into an environment as a living > agent the complexity increases possibly even dramatically. Hypothetically > is my identical twin who is a world renowned polymath verses me a lifelong > alcoholic hermit is our k-complexity the same? There is a difference from > genomic to present state complexity after flushing out the "code" into a > full organism and performing computational operation with the environment. > > I think 8M might be way too small even with junk in the genome. And the > power of microbial intelligence is based on inter-agent communication, > self-organization, cooperation, collective/swarm intelligence, ability to > mutate, etc.. So how does the complexity scale as bacterial population > scales? > http://en.wikipedia.org/wiki/Microbial_intelligence > > Then looking at Cyc's complexity it's difficult to judge since the pure > data value is not as important as it's interrelational operability in human > interactions so a purely k-complexity view doesn't reflect informational > value perhaps... > > John > > > > > ------------------------------------------- > AGI > Archives: https://www.listbox.com/member/archive/303/=now > RSS Feed: https://www.listbox.com/member/archive/rss/303/23050605-2da819ff > 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
