Todor, The FIRST step in simulation is desk-checking. Try doing yourself what you wish a machine to eventually do. THAT was the goal of my reverse Turing test - something to shed light on potential solutions without writing a line of come.
Continuing... On Mon, Oct 14, 2013 at 9:40 PM, Todor Arnaudov <[email protected]> wrote: > Hi Steve, > > That was a concise functional statement that emphasizes the > generality/versatility point and that the problem we are solving is much > different, it could do anything that humans could, or in particular - what > I myself do understand and practise, as a universal artist and polymath. > > I can be very specific in details of workable and simple AGI applications, > for example I have a prototype/project that's related to the DARPA funded > "CALO", mine is supposed to be more general and versatile, as I myself need > it for various kinds of intellectual and cognitive activities, so it's > exactly about designing, developing and creating anything, the first seeds > were conceived back in 2007. > > > >ANYONE who had ANY real interest in building a workable AGI would start > out with simulations, to see where >the challenges were. > > How else one could start??? This is the default way, there's actually no > other way to do it, the design of the AGI includes constant simulations - > it is a simulation, as well as its education, development etc., you don't > breed it in the body of The Terminator or Skynet. > Oops, there you go wanting to write code, before the analysis is even done, let alone the design. > > However you couldn't really see how exactly it will affect humans and more > specifically the humanity, it would be always only guesses, because there > are so many agents and armies and forces that could do anything. > > The general challenges, I agree, are in humans, for example the fact that > Economy will have to change, but the ones who are now rich and powerful may > be unlikely to agree, or also that the machine will show that humans are > not that smart, and humans won't like the machines, I'm afraid that there > could be a situation like in the "Planet of the Apes" (the original one), > where we will be the apes... :) > I have posted about several issues where the human race is destroying itself, yet would turn and fight any power stopping this self destruction. I have received absolutely NO objections to these arguments. The readers here agree that the practices are self destructive, AND that they would fight to keep these practices in place. > We might be stronger and rule the AGI and use it only as a slave, make her > a "lobotomy" etc., and in that case the humans will be the morally inferior > entities - that's also a point to be considered, because humans > inclinations toward slavery are since the beginning of time, and the > romantic Asimov's "laws of Robotics" are also about slavery and breeding > "smart dogs". > Morals do NOT withstand close scrutiny. If they did, they could be coded into principles that would be broadly agreed upon. > > As of the philosophical questions, I've been investigating them > theoretically since so long back in time, especially relative to my age, I > wrote a novel about that when I was 18, I've written too much about it > already and generally it seems exhausted - however you're right that you > rarely may find someone appropriate to discuss on. > > This essay for example was answering questions from Oxford University > about AGI, for example: > > *Monday, February 20, 2012 > Philosophical and Interdisciplinary Discussion on General Intelligence, > AGI and Superintelligence Safety and Human Moral | Cognitive Origins of the > Concepts of Human Soul and its Immortality | Free Will and How it > Originates Cognitively | Animate Being and Soul and the Cognitive Reason > for the Believe that "Thinking Machines can't have a Soul and > Consciousness" | Technology Making us more Humane | The Egoism of Humanity > | And more* > > http://artificial-mind.blogspot.com/2012/02/philosophical-and-interdisciplinary.html > > > Or this novel (but it's in Bulgarian: > http://eim.twenkid.com/old/3/26/istinata11.htm > > > One of the questions: > > "Oxford: Self-modifying systems: What can be proved about the performance > and capabilities of different kinds of recursively self-modifying programs? > Can a framework be developed in which demonstrably safe, recursively > self-improving AIs could be constructed, with stable and > human-friendly goal systems? " > > In fact humans are the "bad guys", they are PROVEN killers and maniacs, > they are proven to have as leaders people who are not the more intelligent > - usually the most intelligent people are not aggressive, at least not > physically aggressive. > Even Herman Kahn, originator of Mutually Assured Destruction, wrote about the VALUE of vengeance. > > Humans have done all the killing, they are greedy, they destroy, > ... and the problem with this is?!!! > and yes, they do not because of their intelligence, but because of their > LACK of intelligence, and because of their primitive brain areas, their > fears, lust, dopamine-short cuts, their difficulties in communicating etc. > nonsenses. > > --- How could one ensure the safety of humans and human intelligence? How > could one control humans and did we succeed in the history of humanity to > prevent wars or genocides? > We are who we are BECAUSE of wars and genocides. > > It started to become possible to apply some control after there were > appropriate technologies. > Real, "natural" humans are beasts, technologies make humans to be the > humans in their own image about themselves. > > I believe that this will ultimately prove to be self-destructive. Our economy is winding down, the population is so depressed that the birth rate is stabilizing, etc. We apparently need a really major war. Of course there ARE alternatives, but so far I have been unable to find anyone to discuss such issues with. > > Steve> BTW, it is my belief that 99% of what people expect of AGIs is > within the reach of human skills. The REAL challenges are timidity, > conformity, greed, etc., and NOT intelligence. > > It woyld be within the reach of *humanity* skills - of the best of the > humans, picked one-by-one - with all the technology developed, but human > individuals often "seize" merits which do not belong to them. What one > ingenious theoretical physicist or mathematician could see from the data is > not what "a human can do", it's what an ingenious human, who has talents > appropriate for this field, could do. > > As of those behavioral issues: *they are issues of the humans, the > problem is with the greedy, violent,* *frightened, suffering pains > humans, especially the ones who have powers and control weapons etc., who > may want to use AGI for destruction. > * > Yes, technology defeats the democracy of war. Wars should be fought with sticks and stones. > > Something that I try to defend in the essay: > > You don't need a universal thinking machine in order to destroy and kill. > The big humanoid terminators T-600 in the movie for example are nonsense. > It doesn't have to look like a human at all, and it could use much more > deadly and effective weapons than using two hands to hold a gun, aiming > visually, shooting etc. > > British ship guns > http://www.youtube.com/watch?v=BwZ3a_Kie8w&feature=player_detailpage#t=72 > > Russian ship guns > http://www.youtube.com/watch?v=KOlVcRns5dQ > > I remember a radio interview in 2001, regarding the decoding of human DNA > and that the information was published on the Internet. A journalist has > asked a scientist: > > - Wasn't that too dangerous, terrorist could develop new dangerous > chemical weapons? > > The scientists answered, smiling ~ : > > - No, it's unlikely, there are enough natural toxins which are easy to > collect... Why dealing with genomics etc. > > Another memory in this context. According to a show on some educational > channel, after the end of WWII there was a conspiracy to poison the water > of the cities in Germany, as a revenge to the Germans. They has been aiming > to kill 6 millions of Germans, and they could have done it. Hopefully a > wise man who they trusted asked them: will that return the dead ones back? > > So you should be afraid of humans, as Seneka has said ~ even the lowest > slave could take away your life. > A maniac could stab you in the back, a drunk driver could hit you on the > road, etc., but the people are rather afraid of Asteroids hitting the > Earth, "The end of the world" etc. bullshit. > > Any system that could divert asteroids could also direct them to selected targets. I fear asrtroid "protection" **FAR** more than I fear asteroids. > The human values, as elaborated more in the essay, are possible to be > applied de facto only with the assistance of technology and eventually > thinking machines. > > Humans without sophisticated technology and societies to provide enough > food and other supplies and developed culture that suppress the real human > nature, are proven beasts who want to rule and kill each other. We divide > in groups and start wars, from the gangs at school bullying the weaker > ones, to the military forces and world wars. > ... and here we are, with all of our technology, deciding to invade Afghanistan and Iraq, discussing the possible invasion of Syria, etc. > > By the way, guys, do you realize that even the "Terminator II" movie is > about how bad the humans are, but people don't get it, even though it's > told even literally in the conclusion. > > Terminator II is about HUMANS being assholes and a machine that learns the > value of human life and tries to save humans for themselves and it > "self-terminates" to protect them. As even the ultimate thinking machines > hater, who lost her beloved one by a thinking machine and whose son and > herself was about to be killed by a thinking machine concludes, after > becoming emotionally attached to the "killing machine": > > *"Because if a machine, a Terminator, can learn the value of human life, > maybe we can too."* > Then, can we learn the value of human DEATH?!!! You seek to stop human evolution. Wouldn't it be better to optimize the process? > > In the setting of the movie and the reality: humans had Cold War and USA > vs USSR and Nuclear weapons facing each other, humans were and are military > maniacs due to their fears, greed and miscommunication. > > The ones who are afraid of the thinking machines should actually be more > afraid of the crazy politicians and crazy generals around the world, the > number of countries with nuclear weapons is steadily growing, and the ones > who own such weapons are increasingly dangerous. > > Sure, there is craziness at many levels, any one of which can destroy our world. Why seek to feed this with AGIs, before we have stabilized our society? I see **NO** prospect for AGIs to be anything but destructive in our presently unstable world. You seek to "patch" an AGI to work well, despite its crazy overlords. So far, no one has made even a flimsy case that this is even possible, given the ease that such patches could be removed. What isn't visible is that you are I are among the crazies, because we believe that some things are good that will almost certainly prove destructive if implemented, e.g. stopping the evolutionary process, preserving social practices that keep lifespans short, etc. There can be no hope for AGIs until their creators can at least see their own craziness. Steve > On Mon, Oct 14, 2013 at 10:43 AM, Todor Arnaudov <[email protected]>wrote: > >> >> From: >> http://www.designweek.co.uk/news/james-dyson-award-2013-international-shortlist-announced/3037334.article >> >> "....Asked simply to ‘design something that solves a problem’, design >> students and recent graduates are competing for a £30,000 international >> winners prize that will also see £10,000 go to their institution. Regional >> winners have already been identified......" >> >> That's too easy, the true AGI, or Self Improving General Intelligence, or >> Versatile Limitless Self Improver is towards something more ambitious: >> >> >> *Something that solves *ALL* problems and that will *be capable to >> design* and complete them if given the physical resources, and also any >> other conceivable projects, and will allow human designers, artists, >> engineers - anyone - to boost their creativity and productivity to super >> human levels.* >> > > >> >> >> -- >> ===* Todor "Tosh" Arnaudov ===* >> * >> .... Twenkid Research:* http://research.twenkid.com >> >> .... *Author of the world's first University courses in AGI (2010, 2011) >> *: >> http://artificial-mind.blogspot.com/2010/04/universal-artificial-intelligence.html >> >> *.... Todor Arnaudov's Researches Blog**: * >> http://artificial-mind.blogspot.com >> * >> * >> > > > > -- > ===* Todor "Tosh" Arnaudov ===* > * > .... Twenkid Research:* http://research.twenkid.com > > .... *Author of the world's first University courses in AGI (2010, 2011)* > : > http://artificial-mind.blogspot.com/2010/04/universal-artificial-intelligence.html > > *.... Todor Arnaudov's Researches Blog**: * > http://artificial-mind.blogspot.com > * > and an > * > *AGI* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/10443978-6f4c28ac> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > -- Full employment can be had with the stoke of a pen. Simply institute a six hour workday. That will easily create enough new jobs to bring back full employment. ------------------------------------------- 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
