On Tue, Dec 25, 2012 at 3:30 PM, Ben Goertzel <[email protected]> wrote: Once we can make a compelling demonstration that genuinely showcases synergetic interaction of various cognitive processes in OpenCog, we will do so; and that will indeed be quite satisfying. The fact that you, or other skeptics, think this is taking longer than it should, isn't particularly important to me.... You understand neither the underlying concepts nor the practical obstacles...
---------------------------- Why would you make a petty personal comment when you were handling the criticism so well up until then? Let's see: I understand neither the underlying concepts nor the practical obstacles... But the criticism that you were responding to was directed at what I see as your inability to take the results of your own scientific experiments and using them to see if you can solve a few problems which would then make your program work a little bit better than it had. Your responses all seem to be made using a form of denial by excuse. This isn't Ben Goertzel Bashing. I am trying to help you deal with the obvious fact that your approach to AGI hasn't worked yet and therefore there is probably is a major problem that you haven't worked out. And I am trying to get this idea across to other people in the group: You have to build a better model for prediction, experimental testing and then you have to utilize the results of your experiments with the courage to learn from your mistakes. You make the claim that you will be able to show a compelling demonstration. Ok, when is this going to happen? What will it show? If you aren't able to make a deadline then what were you able to show in your demonstration? I will keep this message as a draft and post it again next year and every year until you make your compelling demonstration. Jim Bromer On Tue, Dec 25, 2012 at 3:30 PM, Ben Goertzel <[email protected]> wrote: > > Jim, > > >> But does your faith include the possibility that your specific >> ideas might be wrong? >> > > Yes, I accept the possibility that some or all of my ideas might be wrong. > > >> What we need is some way to evaluate the ideas that are being floated and >> tried. The only evaluation that we have is to look at the promotions and >> predictions that researchers are making and compare them with the results >> that they are able to produce. We can also look at how people's agi models >> have improved over the years. >> >> Can you make a better AGI program than your second life dog? If you >> can't, even after all these years, then your prognostications are wrong >> (and maybe you should try another approach). >> > > I'm more interested in expending effort improving the basic infrastructure > and algorithms of OpenCog, in ways that I think support fundamental > progress toward AGI, than in making shiny demos to try to convince skeptics > that the approach is viable.... The "second life dog" demo you refer to > showcased only one OpenCog learning algorithm (MOSES, used in a certain > way) and wasn't intended as some sort of summation of the capability of the > underlying AI approach... > > I think OpenCog can work for AGI. I don't claim it's the only workable > approach, and I would switch to another approach if I saw something that > looked more promising to me. > > Once we can make a compelling demonstration that genuinely showcases > synergetic interaction of various cognitive processes in OpenCog, we will > do so; and that will indeed be quite satisfying. The fact that you, or > other skeptics, think this is taking longer than it should, isn't > particularly important to me.... You understand neither the underlying > concepts nor the practical obstacles... > > Deep learning is getting a lot of attention lately, but I'm not yet > convinced that current deep learning algorithms/architectures have > significant potential beyond the domain of machine perception. (The > general concept of "Deep learning" is surely universal; but there's no > shortage of general concepts with universal applicability to cognition...) > > -- Ben G > > > > > > > *AGI* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/10561250-470149cf> | > 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
