Folks who are obsessed with the historical failure of AI to achieve AGI, really amuse me....
I mean: the field has only been around 60 years. This is minimal in the scope of human history; and the available hardware is still very weak in important respects compared to a mouse brain, let alone a human brain... The failure of AI tech to achieve AGI so far, is certainly not compelling evidence that all the underlying concepts are basically wrong ;-p -- ben G On Sun, Dec 30, 2012 at 1:34 PM, Mike Tintner <[email protected]>wrote: > 1) Ben: I, personally, think it will be best to connect AGI systems > to robots in order to achieve human-like general intelligence.... In such > a system, the logic and math underlying the AGI's algorithms will be used > to interpret the data from the robot's sensors, and to control the data fed > to the robot's actuators.... > > All very well in broad principle. But the *evidence* is that > robots/computers cannot yet produce any AGI-worthy percepts or concepts of > the world whatsoever – cannot really make any AGI-worthy sensory contact > with the world – and are as vast a distance away from understanding even > the simplest of real world scenes, let alone individual objects, as they > are from understanding variations on the simplest of two/three word > sentences (as I have explained elsewhere). This has proved an > extraordinarily intractable set of problems. > > Not knowing where the solutions lie – what form successful percepts and > concepts of the world will take, - you are nevertheless proceeding on the > basis that logic and maths will provide the solution, and are central to an > AGI architecture, despite their extreme and consistent lack of success so > far. Far from being a troll in this area, I repeatedly try to engage with > the problem here – with the forms that successful perceptualisation and > conceptualisation must take. You simply aren’t interested in any solution > that isn’t logicomathematical – isn’t within your current competency – and > don’t have any ideas to offer about how logic & maths or anything else will > directly solve these problems. > > I would say that you are the true AGI troll – by not engaging directly > with the real problems,you are holding the field back. > > 2) Ditto re your rocket analogy. You are actually a rocket builder who has > defied every principle of serious creativity by building an extremely > complex machine, without the slightest evidence or proof of concept, that > your machine is equipped for flight – for its principal function – in this > case for AGI. You have designed a lot of engine/architecture, so to speak, > but you don’t have a take-off mechanism period – and cannot explain in any > way how your machine will achieve takeoff or airborne thrust – how it will > be either general or creative. Wow. > > Again I would say you are the real troll – this is not the way to go about > a serious project, and absolutely the wrong example to set - you are > again characteristically not engaging with the main problem first – which > is what must be done. And many – many AGI-ers, - not just me, have told > you this. (Partly in your defence, this is a field-wide phenomenon, but it > is still terrible). > > 3) Because I am engaging with the problems, I will in the not too distant > future, explain the exact nature of real world reasoning – and how it is > totally different from logicomathematical induction/deduction etc. (To talk > of it being “imaginative reasoning” is to identify the medium of its > reasoning, but not the main ways in which it differs as a form of > inference). > > > > > > > > > *From:* Ben Goertzel <[email protected]> > *Sent:* Sunday, December 30, 2012 4:56 PM > *To:* AGI <[email protected]> > *Subject:* Re: [agi] Why Logic & Maths Have Sweet FA to do with Real > world reasoning > > > > On Sun, Dec 30, 2012 at 5:31 AM, Mike Tintner <[email protected]>wrote: > >> You & others think that an AGI computer is going to be wise about the >> world by sifting through textual documents, and working out logically >> whether “JOHN LOVES MARY”. > > > > > You're battling a straw man here, hmm? I've already been explicit that > I, personally, think it will be best to connect AGI systems to robots in > order to achieve human-like general intelligence.... In such a system, the > logic and math underlying the AGI's algorithms will be used to interpret > the data from the robot's sensors, and to control the data fed to the > robot's actuators.... > > Of course, the proof of the pudding will be in whether the completed > system manages to learn how to control the robot intelligently... > > I imagine you as a troll watching a team of engineers and scientists build > a rocket. You, the troll, deny the theory of rocketry has any relevance > to building real rockets, because that theory uses math, and the world is > after all not obviously made of math. Anyway they didn't teach you much > about math back in troll school. Meanwhile, the engineers and scientists > are about 25% done with building their rocket. And you're wasting your > time, and causing annoyance, standing there pointing your trolly little > finger at them, screeching "Hahaha!! You idiots, can't you see that your > supposedly 25% complete rocket can't fly 25% of the way to the moon?? > Don't you understand that if you want anyone to believe your completed > rocket will fly to the moon, you need to show success according to > quantifiable metrics of incremental progress; you need to show real > evidence!!!" > > Eventually, when the rocket blasts off, the troll brushes the ashes off > his beard and somehow convinces himself it was just an illusion... > > Duuuhhh... > > -- Ben G > > *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/303/212726-deec6279> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > -- Ben Goertzel, PhD http://goertzel.org "My humanity is a constant self-overcoming" -- Friedrich Nietzsche ------------------------------------------- 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
