EFTA02597271

From: on behalf of Ben Goertzel < Sent: Monday, December 22, 2014 7:55 AM To: Jeffrey Epstein Subject: Re: "Intelligent Virtual Animal" research proposal > Hi, Ah, and one more thing regarding my financial particulars. As each year, the OpenCog HK project requires a "corporate contribution" (10% corporate contribution is required to secure the 90% government funding, under this kind of research grant). This year the 10% corporate contribution amounts to roughly US$25K. It's supposed to be given to them by the end of 2014, but if it's slightly late that's probably OK. As usual, I don't have the money and have to dig it up somewhere (which has been from you every year so far!!).... If you choose to fund the overall project that Cosmo, Joscha and myself have suggested, this $25K could come out of that budget, and we could allocate one of the OpenCog HK team members to participate in the project. On the other hand, if you don't choose to fund that overall project, obviously the $25K would still be very valuable. (As you will recall it can be done either as a donation to a charitable organization, Humanity+, or as a payment to a company ... your lawyer has done it both ways, in different years..) The current HK grant funding is for natural language dialogue, and I believe that by the end of 2015 we can get an interesting NL dialogue system created using OpenCog -- and connected to David Hanson's robots, which will be a very effective demo. I don't think this is as scientifically profound as the virtual-animal project, however, it certainly will be a cool demo of OpenCog actually doing something (anything with David's robots is cool-looking!), and will be a genuine advance in NL dialogue systems. As opposed to Sid or Cortana, the dialogue will not be as broadly practical -- but it will be based on underlying reasoning and understanding. As you know, since I am associated with the sponsorship of the OpenCog HK grant at HK Poly U, I can't be funded under that grant. However, I am very very eager to be able to spend my own time on AGI R&D once again (similar to how Joscha is now doing).... Funding of the virtual animal proposal would allow that, which would be awesome... I understand the progress on OpenCog and my related research has not been as excitingly visibly demonstrable as you would have liked... However, a lot of great stuff has been done under the hood, and huge conceptual progress as well as code progress has been made. I think the next two years can be a breakthrough period for the project, and for me. Some things need to percolate for a while, then they come to a boil !!! EFTA_R1_01782710 EFTA02597271 If you have more questions, concerns etc. or just want to discuss out of general interest, let's set a time and have a chat - - I'm generally available (though I guess you may not be, given the holiday season...) thanks Ben On Mon, Dec 22, 2014 at 3:41 PM, Ben Goertzel <-> wrote: > Hi Jeffrey, > After a bunch of discussion and collaboration, here is a proposal from > myself, Cosmo and Joscha, describing some collaborative research work > we would love to do over the next 18 months. > The basic idea is to "start simple" (while building on some of the > powerful tools that Joscha and I have built over the last years), and > aim to scientifically validate our approach (a subset of OpenCog, > including OpenPsi which is modeled on Joscha' Psi/MicroPsi model and > which Joscha would like to help evolve further) in the context of a > virtual animal.... Specifically we would like to implement a bunch > of parrot-oriented animal intelligence tests in a virtual world, and > have OpenCog control a "virtual psittacene bird" that reasons and > learns about its virtual environment and (without hacking specifically > for the tests) performs on these tests similarly to a real parrot like > Irene Pepperberg's Alex. > Due partly to Joscha's influence, this is more science-oriented and > less engineering-oriented than my prior proposals to you, but > nevertheless, in the course of doing this experimental intelligence > evaluation, lots of valuable engineering and testing will get done... > We started out talking about doing something involving experiential > language learning (which would leverage the OpenCog NLP work currently > going on in Hong Kong). However, on reflection, it seemed that if we > wanted to do an experiential language learning project in a really > solid way, it would be best to •firsts get "parrot level language and > animal behavior" truly compellingly, demonstratedly solved -- and then > move on to more sophisticated language. (Of course, human toddler > language is not evolutionarily descended from parrot language, but if > one looks at the issues more abstractly one can see a progression > there.).... So if this project goes forward, I'd say a natural next > step would be to follow this up with a "virtual toddlers in a virtual > world learning language" project. But, first things first.... > When you have time, perhaps the four of us could do a Skype call (or > good old telephone conference call) to discuss? > Thanks a lot, 2 EFTA_R1_01782711 EFTA02597272 > Ben > -- > Ben Goertzel, PhD > http://goertzel.org > "The reasonable man adapts himself to the world: the unreasonable one > persists in trying to adapt the world to himself. Therefore all > progress depends on the unreasonable man." -- George Bernard Shaw Ben Goertzel, PhD http://goertzel.org "The reasonable man adapts himself to the world: the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." -- George Bernard Shaw <?xml version="1.0" encoding="UTF- 8"?> <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> <plist version="1.0"> <dict> <key>conversation-idgkey> <integer>309789</integer> <key>date-last-viewed</key> <integer>0</integer> <key>date-received</key> <integer>1419234887</integer> <key>flags</key> <integer>8590195717</integer> <key>gmail-label-ids</key> <array> <integer>2</integer> </array> <key>remote-id</key> <string>465504</string> </dict> </plist> 3 EFTA_R1_01782712 EFTA02597273