EFTA00773916.pdf

From: on behalf of Ben Goertzel <Milla>1 I To: Jeffrey Epstein<[email protected]> Subject: Re: Continuing our conversation and: AGI Roadmapping Workshop Date: Tue, 25 Aug 2009 17:35:11 +0000 Also, I thought a bit more about specifically what I could accomplish this fall if I obtained research funding (from you or somebody else). I could wrap up a book that is under contract to World Scientific Press, called "Real World Reasoning" ... a sequel to my 2008 book "Probabilistic Logic Networks" (Springer), specifically addressing the math and CS of reasoning about time and space. This is an adapted version of a report I was paid to write for IARPA (a government agency) early this year, by the way. Then I could write a couple papers for the AI Journal on the connections between my AI design and Marcus Nutter's abstract theory of general intelligence. Tjhese already exist in rough draft form but still, polishing math-y stuff for publication takes time and effort. I think that would take me thru the end of October. Then in Nov and Dec I could **finally** finish up a draft of my book explaining my design for an AGI, and send it to some colleagues for preliminary feedback. Basically what I want to do is to re-cast the material in this book-length wikibook http://opencog.org/wiki/0penCogPrime:WikiBook in terms of the "mathematical theory of intelligence" stuff. Thus making a systematic scholarly argument for my approach to AGI. Which I have wanted to do for a long time, but the grounding of the concrete AGI stuff in Nutter's theory has filled in a missing link that makes me understand how to make this a Great Book now, instead of just an interesting one... While doing this writing, I can also continue to lead the OpenCog open-source AI effort, which has some participants independently of my ability to finance work (though of course it will be more active if Novamente, you, Dan or someone comes through with funding for the work) Anyway if you have time I'd love to chat about the above F2F or via phone, whenever you have time... thanks much ben On Sat, Aug 22, 2009 at 8:41 PM, Ben Goertzel > wrote: > Hi Jeffrey, • > When we talked last, you asked me to email you this weekend so I'm doing so. • > If you've thought about what we discussed a little more and would like > to continue our conversation, you can try me anytime on my cell at 240 > 505 6518, or else email me and we can set up a specific time to talk. > Of course I'm still very eager to continue our conversation, as I > would really like to be able to spend this farll writing up my AI > research from the last 6 months rather than hunting for and then > working in some brain-dead government "R&D" job ;-p > Like I said before, strategically I think things look OK for me... the > US economy seems to be rebooting so that Novamente may start > generating more revenue again; and independently of that, I'm resolved > to apply for academic jobs for Sept 2010 (unless you or someone else > chooses to seriously fund an AGI project very soon!). But even so, in EFTA00773916 > the immediate term, this is an extremely difficult moment for me > financially (I don't want to bore you with the details ... but, I'm > staring at my 2008 tax bill which is rather frightening ;-( ... like I > said on the phone it's not a tragic situation because I have family I > could borrow money from, but it's still pretty unpleasant...) > I mentioned the Singularity Summit -- I should also mention another > event in October which is more narrowly focused on AGI, the "AGI > Roadmap Workshop." This is an attempt to gather together 10 or so > real AGI researchers to help plot a common course together for the > field. This will be in Knoxville oct 24-25 ... I realize you're > probably too busy, but you're certainly invited to attend if you have > the time and interest! See • > http://www.ece.utk.edu/-itamar/AGI_Roadmap.html • > for more information. > Between the roadmapping effort, my theoretical work earlier this year, • > and the consistent engineering progress my team has been making > (hooking together our language and virtual-agent learning engines ... > and the next step is to link our probabilistic reasoning engine into > that unified system), I actually feel like we're moving steadily and > meaningfully toward advanced AGI. But it's still research -- it's not > the sort of thing that's going to make anyone rich this year; and it's > also swimming against the current of the US funding sources (though > not the Chinese funding sources, frustratingly ;p) • > Anyway I look forward to talking to you again... your ongoing support > is much appreciated, both financial and also conceptually as "another > rare person who actually understands the value and potential of > advanced AGI" > thanks > Ben > Ben Goertzel, PhD • > CEO, Novamente LLC and Biomind LLC > Director of Research, SIAI • > "Truth is a pathless land" -- Jiddu Krishnamurti Ben Goertzel, PhD CEO, Novamente LLC and Biomind LLC Director of Research, SIAI "Truth is a pathless land" -- Jiddu Krishnamurti EFTA00773917