My current understanding of the problem is as follows: Even humans have to make decisions based on some type of logical/algorithmic thinking. People may seemingly make random choices but they are still constrained by rigid logical parameters that sit above every decision. If they are not, those individuals will not exist on this earth for very long. Since logic is important to biological procreation, nature has built a lot of components into human bodies that automatically generate clear signals of GOOD and BAD, or 1 and 0, if you prefer that notation. In general we wish to maximize the GOOD signals and minimize the BAD signals, within the fairly complex framework of human society.
The brain is good at exploring many possibilities at once due to the parallel nature of the brain's neural network. But a computer is better at taking a couple pieces of information (such as current state and goal) and refining it to the most meaningful action that can be derived from it. This is why computers can make incredibly complex and rational decisions that even the most brilliant human mind would struggle with (in a particular domain, such as chess or networking), but suffer from an extremely narrow scope when compared to humans. The solution is not to abandon the fundamental computational framework that the Turing Machine proposed, but to apply it more rigorously to the computational resources that are available. I think people like Jeff Hawkins are on the right track with regards to creating machines that can would be able make intelligent decisions in a more "human-like" manner. On Mon, May 13, 2013 at 8:09 AM, Mike Tintner <[email protected]>wrote: > My impression is that the TM2 did *start* to have an impression on > minds here – though almost nothing can remove Ben et al’s mental > straitjacket – because you can begin to SEE what I am talking about. > > That’s why the TM paper was so great – you could SEE how it worked – you > weren’t just dealing with abstract maths. > > So let’s show you visually the difference between a TM/algo machine and a > TM2 creative machine from the outside – in terms of what they DO. > > You guys and the whole of science in fact are basically mechanics – you > think about how the machine/car works – you don’t look at the big picture > of what it’s supposed to do and how and where you can drive the car. You > take all that for granted – when, for AGI, that’s the starting-point. > > Here’s a picture of a TM/algo: > > http://www6.pcmag.com/media/images/248747-tug-s-whiskers.jpg > > it’s basically a journeyrobot that follows a ROUTE MAP. Like a hospital > robot. A machine that has a route map which tells it exactly how to get > through its maze, and in effect shows it the whole maze before it starts: > > . > http://2.bp.blogspot.com/-O8hv9dAmGoI/Tt8dIBrUN9I/AAAAAAAAAL0/2IuWDRlEUC0/s1600/maze%2B%25281%2529.gif > > A TM is a machine that moves along precharted lines through precharted > territory. > > Like, also, a warehouse robot > > I > http://cache.boston.com/universal/site_graphics/blogs/bigpicture/robots_03_04/r25_14061299.jpg > > Nice work if you can get it – but there’s one limitation – a TM/algo robot > can’t go anywhere else but that hospital/warehouse, that one charted > territory and along those charted lines.It can’t go out and into the wide > world It can’t go anywhere uncharted. > > Now here’s a very different TM2 – a true AGI - > > > http://cache.boston.com/universal/site_graphics/blogs/bigpicture/robots_03_04/r18_0170-006.jpg > > it’s an EXPLORER robot – only not the present telecontrolled limited > reality – but what an explorer robot should be – a machine designed to > explore foreign territory like a foreign planet – WITHOUT A ROUTE MAP- > WITHOUT humans directing it.. The opposite of a Tm/ALGO. > > So don’t see it so much as exploring Mars – because the Mars landscape is > relatively open and easy to see. > > Think of an explorer that can explore a jungle > > > http://www.google.com/imgres?safe=off&biw=1644&bih=748&tbm=isch&tbnid=M6xPtj5iOUp6MM:&imgrefurl=http://www.wallpaperdev.com/wallpaper/1024x768/new-african-animals-you-are-viewing-the-jungle-scenes-named-palms-it-has-205911-28819.html&docid=sCDBtV7ZlR67UM&imgurl=http://wallpaperdev.com/stock/new-african-animals-you-are-viewing-the-jungle-scenes-named-palms-it-has-205911.jpg&w=1024&h=768&ei=4MqQUeiaPJHu0gXFyYGYCg&zoom=1&ved=1t:3588,r:12,s:0,i:185 > > WITHOUT being able to see the whole jungle before it starts > > Or that can explore – and clean up - the “jungle” of a new room: > > > http://2.bp.blogspot.com/_lBXQVHyQ1VA/S7FXEN2icBI/AAAAAAAAAEY/CqTTaIsyFuA/s1600/messy-room-03.jpg > > Or explore the jungle of the internet: > > > https://www.google.co.uk/#output=search&sclient=psy-ab&q=jungle&oq=jungle&gs_l=hp.3..0l4.1830.2652.0.4007.6.5.0.1.1.0.97.359.5.5.0...0.0...1c.1.12.psy-ab.ttVPw8hO1x4&pbx=1&bav=on.2,or.r_cp.r_qf.&bvm=bv.46340616,d.d2k&fp=40f391f6da3e1fea&biw=1644&bih=748 > > (Mars is actually also a jungle but of a less obvious kind) > > This is a machine that can truly explore a new territory independently – > move along uncharted lines through uncharted territory WITHOUT a route map. > The complete opposite of a TM/algo that can only go where it’s told to go. > > A machine that can endlessly explore new fields without having to be given > a step-by-step itinerary – a machine like you and every other living > creature – would be kinda useful. > > Now if you try the algo/route map approach to AGI – to building a robot > that by definition doesn’t and can’t have a route map – guess what’s going > to happen?. Your machine isn’t going to go anywhere. And that’s the > insanity of present AGI . It’s doing the complete opposite of what it > should be doing. > > Oh but wait a minute – Turing is God, and God said “Thou shalt not and > cannot go anywhere without having a route map beforehand.” > > Er, sorry about this, but God’s wrong. You don’t have to have a route map > to go places and explore the world. In fact, you can’t have a route map for > unmapped, uncharted territory. It’s a physical and computational > impossibility. > > (It helps to look at not just how machines work but what they do – from > the outside as well as the inside).. > > ** > ** > ** > ** > ** > ** > ** > ** > ** > ** > *w* > > > > > > > > > > > > > > > *AGI* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/6952829-59a2eca5> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > *AGI* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/12578217-f409cecc> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <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
