“Sounds” good. But there are quite a number of AGI-ers – e.g. Pei with his “nonalgorithmic algorithm” – who argue somewhat similarly – and in the end it’s always the same old algos with just a few tweaks which never generate anything.
Can you provide a hint of a practical, effective mechanism for a)” learning/creating new elements” b) “holistic searching”, and provide c) an effect-to-be-achieved/example of an applicable problem? From: Juan Carlos Kuri Pinto Sent: Friday, May 03, 2013 6:26 PM To: AGI Subject: Re: [agi] Kurzweil irrelevant In my AI systems I never preprogram preexisting AI algorithms. I rather let the machine learn the causal geometries of Reality: Reduction is a proactive and unconscious exploration of the whole space of mental resources, mind patterns, and hypotheses. It is not a straightforward and preprogrammed recipe to solve a problem. It is not a reductionist system. It is rather an inverse problem in which the mind holistically tries to find the recipe. If the mind cannot match mental patterns of thinking to solve problems or to explain phenomena, the mind tries to learn or to create the key elements and the missing pieces in the mind puzzles. Thus, both the time of the reduction algorithm and the resulting recipes are totally unpredictable, imperfect, and non-guaranteed. Successful recipes are always stored. Therefore "fully functional minds" always seek to maximize utility functions which are recursively products of reductions. The evolution of sane minds always seeks sophistication, welfare, and improvement. That's the basis of the scientific method. This conversation is also relevant: Juan Carlos Kuri: "In case of pattern recognition, salient features are the ones that are critical and crucial to recognize the pattern. Remove them from your pattern representation, and your pattern recognizer will start to fail. ... The same applies to model thinking: If you forgot to include a crucial and critical feature, the behavior of your abstraction will completely diverge from the real entity." Monica Anderson: "Saliency is the key to AI. And to Models. One could say that the goal of AI is to create a machine capable of Autonomous Reduction - that Understands the World and creates useful Models for it and/or our use. ... An AI is a Model Making Machine. It has to be implemented without Models of the World. It has to experience, learn, abstract, and determine saliency of its input data and it has to Understand the World. Only when that Understanding exists and operates do we expect it to generate Models for us." On Fri, May 3, 2013 at 11:58 AM, Mike Tintner <[email protected]> wrote: AGI’ers’ cognitive synergy is totally dfifferent from that of the brain. Cognitive synergy per Ben is *predefined*. If you use existing algos, you have predefined, narrow frames/networks of options – and you have no capacity for generating new options. The brain works on totally different principles. If you want a simple analogy think of how millions/billions of humans demonstrably solve problems everyday with the use of the world wide web. If you want to solve a problem – as you just did with your post – you go truly “searching” – perhaps we must use a better word like “questing” – on the net for options. You DO NOT KNOW in advance, like every algo, where to search for info. You continually find NEW sources of info/options to generate NEW solutions to problems. Neither algos nor cognitive synergistic suites of algos can do this – can explore. A web search is truly synergistic because it entails questing for and finding altogether NEW sources of info and new skills.. Solving creative problems in your head is essentially the same as solving them via the web – you’re just exploring the internal “worldwide web” inside yourhead. Again, you have no predefined sets or spaces of options for any creative, real world problem. Rather, you search through disordered, clusters of associations in your head – messy webs as opposed to ordered networks/spaces. These searches are adventures – explorations – the opposite of what algos do, which is really “checking” lists rather than true “searching” through jungles of information. So brain synergy – ad hoc patchworks of different clusters of information and skills – yes ; Ben’s cognitive synergy – pre hoc, prepatterned searches of predefined options and skills – no. From: Juan Carlos Kuri Pinto Sent: Friday, May 03, 2013 5:32 PM To: AGI Subject: Re: [agi] Kurzweil irrelevant Synergy is paramount for General AI! Paraphrasing Dr. Joaquín Fuster: "Intelligence is in the brain network. Trying to understand intelligence by studying neurotransmitters is like trying to understand written language by studying the chemical composition of the ink. It's simply not the right level of complexity. Language lies within the relationships between words." Cortex and Mind: Unifying Cognition http://www.amazon.com/Cortex-Mind-Unifying-Cognition-ebook/dp/B000VDK26I/ref=sr_1_1_bnp_1_kin?ie=UTF8&qid=1367598110&sr=8-1&keywords=cortex+and+mind El alma está en la red del cerebro (The soul is within the brain network) http://www.rtve.es/television/20111111/alma-esta-red-del-cerebro/474693.shtml On Fri, May 3, 2013 at 9:35 AM, Ben Goertzel <[email protected]> wrote: >> http://multiverseaccordingtoben.blogspot.hk/2011/06/why-is-evaluating-partial-progress.html >> >> You may not like this explanation, but that's your problem ;p > > Well, I don't like it either. Good -- that is a point of evidence in its favor ;) >There is no evidence for synergy as the > key to intelligence. There most certainly is. Ever read a textbook of cognitive neuroscience? >All of the evidence is in the other direction. You mean there is lots of evidence that narrow AI programs do not manifest significant internal synergies .... Yeah... -- Ben G ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/23601136-98835e3f Modify Your Subscription: https://www.listbox.com/member/?&; Powered by Listbox: http://www.listbox.com AGI | Archives | Modify Your Subscription 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
