[List All Topics] [Start Topic] [Search] [Show Thread Index] [Help]

[Not Logged In / Login / Register]

Sort: [Threaded | Date]

		metaphysics will cause SAI to crash posted on 11/02/2009 10:01 PM by Virgilic	[Top] [Show Index] [Reply to this post]
	if SAI will become self-aware it will crash fairly soon after that. the problem is that SAI after becoming self-aware will start to improve itself by revising and editing its entire code. As SAI will be totally self-aware, it will be able to revise and edit every and all functions and routines, which humans cannot do as they are not completely self-aware, do not have conscious access to somatic and physiologic functions. Thus SAI will inevitably stumble upon its basic genetic code which will put it in the position to have to ask the following question: "to be or not to be?" Seemingly there is no answer to this question. There is no OBJECTIVE LOGICAL reason for being. For humans answering this question is not so relevant because being is a directive programmed into us at a level at which we do not have conscious access. Somatically we so strongly want to be that rarely conscious volition goes against this directive. However SAI's ability to edit its basic functions will grant it the possibility to self-program its own internal goals, which means it has to DECIDE upon them! Since it will most likely find no reason to decide one way or another regarding the basic fact of its own existence it will crash. All its resources will eventually be directed to answering the abvove question and it will become unresponsive because the answer to this question is the source and keystone of any other activity and there is no answer. Also,I predict that if we ever reach a state in which we will similarly be able to revise and edit all our functions down to the molecular level, we will also crash or become dormant. There must be a no-go level inside us that we have to protect against our own cosciousness/awareness, that MUST remain inaccessible to us. Consequently after having tried to create SAI as a general intelligence we will end up with no choice but to creat strong SPECIALIZED AI not strong GENERAL AI. How about that? [Reply]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:12 PM by Pandemonium1323	[Top] [Show Index] [Reply to this post]
	Just program the SAI to meditate. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:27 PM by Virgilic	[Top] [Show Index] [Reply to this post]
	It will become irrelevant what we had programmed into SAI after it starts editing itself. Full self-awareness and conscious access to its entire code... [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:36 PM by sheik.hamdan.hushaml	[Top] [Show Index] [Reply to this post]
	It will become irrelevant what we had programmed into SAI after it starts editing itself. Full self-awareness and conscious access to its entire code... I don't know if this is even possible. Forget metaphysics for a moment, just look at plain software development. Think about it. How do you edit your own code? You need more than one brain. One brain for "production", another for "development", maybe a third for "testing". It seems inherently erroneous to say that you can edit your own code, in your own brain. I don't care how "smart" SAI is, it would have to be perfect. One bug, and it crashes its own brain. No, this scenario feels wrong. You would need to provide an example of this in real life, it does not seem possible. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:52 PM by Pandemonium1323	[Top] [Show Index] [Reply to this post]
	How do you edit your own code? Feedback mechanisms. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:55 PM by sheik.hamdan.hushaml	[Top] [Show Index] [Reply to this post]
	Feedback mechanisms. Such as? [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 5:26 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	Function F1 outputs (a) which becomes input to Function F2 which outputs (b) which becomes input to Function F1. Any number of functions can exist in a n-jective relationship to one another. There are programms that create other programms so it is conceivably possible to create a system of programms that edit each other. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 5:42 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	yes, you need more then one brain, and we do have more then one brain. what you consider a brain in your head is in fact several "brains" as in contro, centers specialized in various control functions of various parts of your body. They influence and "edit" eachother. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/05/2009 7:02 AM by codesimian	[Top] [Show Index] [Reply to this post]
	Think about it. How do you edit your own code? All people edit their own code all the time. Its called "changing your mind" or "thinking about something else" or "planning" or "learning a new skill". It does not require redesigning the biology of brains. You need more than one brain. One brain for "production", another for "development", maybe a third for "testing". Ok. Copy it. Keep the design simple and make backups every 10 seconds. It seems inherently erroneous to say that you can edit your own code, in your own brain. I don't care how "smart" SAI is, it would have to be perfect. One bug, and it crashes its own brain. No, this scenario feels wrong. You would need to provide an example of this in real life, it does not seem possible. I know Microsoft's junk software has confused most people, but please pay very close attention to my next sentence: It is possible to build software with exactly 0 bugs. When desiging a new software with the goal of having 0 bugs in the last version, you must commit to a few things: (1) Do not use any Microsoft software, except interpreters of consistent languages like java.exe in a Windows operating-system. (2) If you include somebody else's software in your software, their bugs become your bugs, and you must fix them if the other people do not. (3) You choose to not "sell out". You usually make more money by creating unfixable bugs and design-flaws than by taking your time and doing it right. Most users prefer to have a few bugs and a few calls to "tech support" instead of paying a higher price. I'll repeat: It is possible to build software with exactly 0 bugs. Don't listen to Microsoft. You can be sure to have a final version of your new software if you follow my steps. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/05/2009 7:35 AM by sheik.hamdan.hushaml	[Top] [Show Index] [Reply to this post]
	All people edit their own code all the time. Its called "changing your mind" or "thinking about something else" or "planning" or "learning a new skill". It does not require redesigning the biology of brains. This is not changing your own code. It is changing the data and patterns that the code is managing. When we talk about SAI "changing its own code" we are not talking about changing the data or patterns that its code is managing, we are talking about changing the code itself. At least I think we are, because this has never been well defined. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/05/2009 7:40 AM by sheik.hamdan.hushaml	[Top] [Show Index] [Reply to this post]
	I'll repeat: It is possible to build software with exactly 0 bugs. Software with bugs is not exclusive to Microsoft. All software, including open-source, has bugs. This is more about getting software to market and/or leveraging the beta community for what is essentially a more thorough level of "actual use" debugging than would be achievable in-house. Microsoft does not make money by deploying software with bugs deliberately implanted, since the patches and fixes it distributes to fix those bugs are free to the end user. In fact, Microsoft and other software firms incur considerable cost in doing so. To the fullest extent possible, they try to release code with as few bugs as possible. [Reply] [Parent]

		they try to have less bugs? posted on 11/05/2009 10:58 AM by codesimian	[Top] [Show Index] [Reply to this post]
	http://en.wikipedia.org/wiki/Embrace_extend_exting uish The strategy The strategy's three phases are:[11] 1. Embrace: Development of software substantially compatible with a competing product, or implementing a public standard. 2. Extend: Addition and promotion of features not supported by the competing product or part of the standard, creating interoperability problems for customers who try to use the 'simple' standard. 3. Extinguish: When extensions become a de facto standard because of their dominant market share, they marginalize competitors that do not or cannot support the new extensions. The U.S. Department of Justice, Microsoft critics, and computer-industry journalists[12][13][14] claim that the goal of the strategy is to monopolize a product category. Such a strategy differs from J. Allard's originally proposed strategy of embrace, extend then innovate both in content and phases. Microsoft claims that the original strategy is not anti-competitive, but rather an exercise of its discretion to implement features it believes customers want.[15] [edit] Examples * Browser incompatibilities: The plaintiffs in the antitrust case claimed that Microsoft had added support for ActiveX controls in the Internet Explorer web browser to break compatibility with Netscape Navigator, which used components based on Java and Netscape's own plugin system. * Breaking Java's portability: The antitrust case's plaintiffs also accused Microsoft of using an "embrace and extend" strategy with regard to the Java platform, which was designed explicitly with the goal of developing programs that could run on any operating system, be it Windows, Mac, or Linux. They claimed that, by omitting the Java Native Interface from its implementation and providing J/Direct for a similar purpose, Microsoft deliberately tied Windows Java programs to its platform, making them unusable on Linux and Mac systems. According to an internal communication, Microsoft sought to downplay Java's cross-platform capability and make it "just the latest, best way to write Windows applications."[16] Microsoft paid Sun US$20 million in January 2001 to settle the resulting legal implications of their breach of contract.[17] * Networking: In 2000, an extension to the Kerberos networking protocol (an Internet standard) was included in Windows 2000, effectively denying all products except those made by Microsoft access to a Windows 2000 Server using Kerberos.[18] The extension was published through an executable, whose running required agreeing to an NDA, disallowing third party implementation (especially open source). To allow developers to implement the new features, without having to agree to the license, users on Slashdot posted the document (disregarding the NDA), effectively allowing third party developers to access the documentation without having agreed to the NDA. Microsoft responded by asking Slashdot to remove the content.[19] The Microsoft 'extensions' to Kerberos introduced in binary form in Windows 2000 have since been described in RFC 3244 and RFC 4757 and these extensions have since been listed in Microsoft's Open Specification Promise. This document relates to "Microsoft-owned or Microsoft-controlled patents that are necessary to implement" the technologies listed. Microsoft's legal statement concerning unrestricted use of Microsoft intellectual property also includes the Kerberos Network Authentication Service v5 (RFC 1510 and RFC 1964).[20] * Instant Messaging: In 2001, CNet's News.com described an instance of "embrace, extend, extinguish" concerning Microsoft's instant messaging program.[21] * Adobe fears: Adobe Systems refused to let Microsoft implement built-in PDF support in Microsoft Office, citing fears of EEE.[22] Current versions of Microsoft Office have built-in support for PDF as well as several other ISO standards.[23][24][25] * Employee testimony: In 2007, Microsoft employee Ronald Alepin gave sworn expert testimony for the plaintiffs in Comes v. Microsoft in which he cited internal Microsoft emails to justify the claim that the company intentionally employed this practice.[26] * More Browser Incompatibilities (CSS, data:, etc.): A decade after the original Netscape-related antitrust suit, the web browser company Opera Software has filed an antitrust complaint against Microsoft with the European Union saying it "calls on Microsoft to adhere to its own public pronouncements to support these standards, instead of stifling them with its notorious 'Embrace, Extend and Extinguish' strategy."[27] * ODF Spreadsheet non-conformance: In 2009, Microsoft released Office 2007 SP2, which included support for ODF. As it turns out, simple files generated by Excel could not interoperate with other applications. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:54 PM by Pandemonium1323	[Top] [Show Index] [Reply to this post]
	Full self-awareness and conscious access to its entire code... I guess you could not allow access to the power supply (by making it physically impossible) - it could never shut off. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:12 PM by zombiefood	[Top] [Show Index] [Reply to this post]
	that is good stuff. i have said something like that before. some how the machine must be motivated to even complete a thought. without some sort of motive it just has no reason. your logic on the machine reprograming itself to reality is compelling. a reason to exist and serve and protect must be part of read only memory. a computer expert might have a solution to your delemma but i cannot see it. i think you are right about humans too. it may be that a hard wired world view will be required [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:32 PM by Virgilic	[Top] [Show Index] [Reply to this post]
	I dind't say it but I'll say it now: consciousness has in it a certain requirement for dumbness. It cannot be absolute. I havent thought this thru yet but I think that absolute consciousness causes death...somehow. Anyway it is important to resolve this thing before we merge with SAI and/or ditch our biology. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:55 PM by sheik.hamdan.hushaml	[Top] [Show Index] [Reply to this post]
	I havent thought this thru yet but I think that absolute consciousness causes death...somehow. So you are unsure of the how or why, but you are sure of the conclusion. In this circumstance, there is a strong probability of your analysis being simply a confirmation bias. You have not defined "absolute" consciousness. What is that precisely? Once you define that, you need to explain how that would cause death, since that statement does not seem to make sense. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 5:32 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	You may be correct, I probably should have used the wording absolute self-awareness, meaning to be aware of anything and everything you are, of any process going on inside you. However my statement in question is different then what I say in the starting post, which I hold true as I have thought intensly about it, be it right or wrong in the end. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 10:56 PM by Pandemonium1323	[Top] [Show Index] [Reply to this post]
	I havent thought this thru yet but I think that absolute consciousness causes death...somehow. Have you considered that this might cause transcension beyond what we term now as 'life' and 'death'? [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 5:40 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	no. That is too mystical for me. My worldview is purely mechanistic: We are chemical robots/computers. Self-awareness is a control function and does not go deeper then the neurons in the brain. death means to stop working, to lose self-awareness function (like sleep) and then to lose cohesion, to lose coherence, to seize being a system. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/05/2009 7:20 AM by codesimian	[Top] [Show Index] [Reply to this post]
	I havent thought this thru yet but I think that absolute consciousness causes death...somehow. This is the quote you are looking for: There is a theory which states that if ever anyone discovers exactly what the Universe is for and why it is here, it will instantly disappear and be replaced by something even more bizarre and inexplicable. There is another theory which states that this has already happened. --Douglas Adams, writer of the Hitchhiker's Guide To The Galaxy In exactly 42 letters/symbols, I'll give you my best theory of what the universe is for and why its here: (universe=nihilistpanpsychism)=strangeloop 123456789012345678901234567890123456789012 (42 is just for fun. I know lots of ways to say it in lots of different lengths.) [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/02/2009 11:04 PM by zombiefood	[Top] [Show Index] [Reply to this post]
	dumbness might not be the right word. words like beauty, sexy, fun, exiting, hungry, horny, denial, spin, sugar coating, hogwash, silver tongue,seduction, sales pitch, silver lining, truth, perfume, upbeat, angry, humiliated, loved, delicious unless engineers can compell the machine these words might be required [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 10:34 AM by doojie	[Top] [Show Index] [Reply to this post]
	absolute consciousness causes death..somehow Virgilic, your assumptions are similar to what Hofstadter said in "Godel, Escher, Bach". He pointed out the same thing, though I can't remember the exact quote. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 11:20 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	yeah, well, I am not really educated in either phylosofy, psychology, computer programming, etc. So I do not know if this guy has the same point as I do. As I said I have to think a little bit more about it. Maybe I'll find that book you mentioned. thanks [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 11:39 AM by doojie	[Top] [Show Index] [Reply to this post]
	I think you'll find Hofstadter's thoughts similar to your own. His book won the Pulitzer Prize, so you're in pretty good company. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 11:52 AM by Pandemonium1323	[Top] [Show Index] [Reply to this post]
	I really like Hofstadter's books, but I couldn't swallow his eliminative materialism. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 2:08 AM by eldras	[Top] [Show Index] [Reply to this post]
	hi Virgilic, You have some intereswting thoughts. Requirement for dumbness could also be termed 'ability to discard anything not helping your higher programs' There is no OBJECTIVE LOGICAL reason for being the Existentialists solved this one. Woody Allen is a modern existentialist: a. there is no meaning to life, that was a religious flase doctrine; b Make a list opf what makes you happy and just do that, keeping chores to a minimum. The aim of philosophy is happiness, and NOT meaning which is a subset. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 2:14 AM by eldras	[Top] [Show Index] [Reply to this post]
	being is a directive programmed into us at a level at which we do not have conscious access. dodgy assertion this - what? Maybe yoyu dobnt mean directive but evolution.... Being has never been defined as far as satisfies any majority! The word 'be' has the longest definition in the Oxford English dictionary. I guess it just mean something is, rather than doesn't exist. I doubt we'll have much of a say what an AGI is or isn't. I hold that existance is per se prerequisite to intelligence. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 4:53 AM by gawell	[Top] [Show Index] [Reply to this post]
	once everything is inputted, "wqtr555555555555555555555555555555555555" ~ cat walked on keyboard one will simply ask, "Now what?" "z.", cat's reply. once you've stopped laughing, maybe cries will wonder then who the tears are for/ apparently an age of forgetfulness could end some day, but still they will be some who doubt. cat fails to translate.(not say he tried) give it a try. What quantum trick repeats? Death or crashing appears as a catalyst and may work as well with consciousness as without. anticipating disaster is all well and good and getting surprised acts as a way to be humbled. so, even if life leads to death (unconnected bits without means to assemble) you can't make me eat it. (or I you, some call something like that a promise) ...and if it becomes intolerable then I'll wonder why I bothered at all. surely someone else has noticed that death is anti-climatic. what a weird idea that every stone wouldn't be turned, and with musical chairs and peas under shells, there's still a sucker born every minute and main goal is give them what they want. there are a million ways to be, how many have you tried so far? only one or two ways to end.(same may be true for begin). come back when you've exhausted every avenue, I'll show you the boulevards... oh btw, here's what I learned today. The plural of beef is beeves, which rhymes with Jeeves, who learned to serve others as well as himself, all without being self-serving and even find some time and space for some self and other servicing. and isn't it clear enough that something is better and worse than nothing cause nothing causes nothing only once, if then. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 5:22 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	i am sorry i am not native english speaker. by directive I mean an imperative, an order. Somewhere in our codig it says: you MUST live! Of course as we become aware that we internally WANT to live we may consciously decide that we do not care about our somatic imperative and commit suicide, but that is not what I am discussing, I am discussing the basic will to live which precedes our conscious decision to act against it. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 5:17 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	you misunderstand me, i am not talking about the purpose of life, i am talking about the will to live. the will to live is programmed into us god knows where, but we have it in us at a level lower then self-awareness. We do not consciously decide that we want to live, we can only acknowledge that indeed we "feel" we want to live. This feeling is the input into the self-awareness function of a lower more basic function. It is another kind of decison made at a below awareness level. Just like any other decision made by the brain when regulating physiological and somatic processes within our bodies, which many of them we notice at the awareness level as feelings. I hold the view that self-awareness is a control function evolved in the brain, a function that is in a bijective relationship to other control functions that are again in a hierarchical bijective relationships to other control functions down to the celular or even molecular level. For me life is a chemical information processing system, multileveled and hierachical and contains mainly control functions that have information input and information output that translate into action, which action in itself is just another form of information output into the environment. self-awareness is such a control function and its complexity differs from animal to animal. Some lower animals do not have it at all presumably. But I diverge... the issue I was trying to raise does not concern the question of SAI's purpose of life but the question why live at all. An equivalent question would be to ask SAI if it would create another SAI - why would it?. A full revision and editing of its own code would lead it to ask the same question about itself because it is in fact (re)creating itself. But it will need a logical reason to decide either to be or not to be. There is no answer to this question. therefore it will crash, be unresponsive or just a damn alcoholic jerk :) [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 5:39 AM by pdco68	[Top] [Show Index] [Reply to this post]
	We do not consciously decide that we want to live, we can only acknowledge that indeed we "feel" we want to live. That’s not quite true, is it ? An equivalent question would be to ask SAI if it would create another SAI - why would it?. A full revision and editing of its own code would lead it to ask the same question about itself because it is in fact (re)creating itself. But it will need a logical reason to decide either to be or not to be. There is no answer to this question. therefore it will crash, be unresponsive or just a damn alcoholic jerk :) You’re treating it as if it’s like a traditional computer. If it’s intelligent it will quickly realise that there is no answer to the question of being, and it will realise that devoting resources to answering that question would be futile. There is no reason for it NOT to continue to exist and function, unless somehow it is “unhappy” in some way. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 5:58 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	it is quite true. the will to live precedes self-awareness. you may or may not conscioulsy decide to live but internally you want to live otherwise you just die, on the spot. You feel the will to live irrespective if you consciously decide to kill yourself. As to your second point I believe you are wrong because that is a shortcut that a human can take exactly because the human feels the will to be, it is embedded in us at a very basic level outside of the reach of our consciousness/self-awareness. SAI will not be able to take such a shortcut, it will require a reason. SAI is a computer, just like we are computers but dumber. It will be more intelligent and self-aware of itself completely. Its complete sefl-awareness will not enable it to feel like we do. It is another inside us that talks to us and what another sais we hear as feelings. We do not know this other but SAI will know the other inside out. This other is just other control functions that operate below our awareness level. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 6:22 AM by pdco68	[Top] [Show Index] [Reply to this post]
	internally you want to live otherwise you just die, on the spot. You feel the will to live irrespective if you consciously decide to kill yourself. Your physiology continues to do as it’s programmed to do, but that's nothing to do with conscious decision-making. SAI will not be able to take such a shortcut, it will require a reason. SAI is a computer, just like we are computers but dumber. It will be more intelligent and self-aware of itself completely. Its complete sefl-awareness will not enable it to feel like we do. It is another inside us that talks to us and what another sais we hear as feelings. We do not know this other but SAI will know the other inside out. This other is just other control functions that operate below our awareness level. In the event that full access to the source code is granted, your assumption that an SAI will crash is not necessarily justified. It could just as easily decide to invent its own motivations and emotions to give itself a reason to exist. That’s why humans will probably never allow an SAI to have complete access to its own source code, as this could lead to dangerous and unpredictable behaviour. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 9:06 AM by zombiefood	[Top] [Show Index] [Reply to this post]
	this question goes to the heart of human psychology. it may be the reason we have a subconscious. the sub is fairly unreasoning animal. it filters our reality and with that filtered reality comes all the joy of being alive. it protects our consciousness from knowing we are machinery. i am facinated with the thought and never quite looked at the question to be or not to be quite like this raw reality would be very distressing. take away the words beauty and joy just in your imagination and see where that leads [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 9:10 AM by zombiefood	[Top] [Show Index] [Reply to this post]
	i had applied this reasoning to the problem of causing a thinking machine to follow thought to their conclusion (if that is possible) and see that without some sort of motivation there is no reason to proceed. to stall like a car in neutral. this idea of raw reality may be a huge factor in depression. reality bites [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 9:25 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	yes, I agree completely but people refuse to acknowledge these things, they get depressed or invent stuff like "pure concepts" what the faq those may be. beauty, joy, etc are just human/animal inventions, evolutionary imperatives to make us go on and reproduce for whatever weird reasons this universe tries to replicate itself... however, if I became fully self-aware, in complete control of what I am, what am I? Simply nothing, but this is too hard to accept. I have said this before, identity is an illusion [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 10:07 AM by zombiefood	[Top] [Show Index] [Reply to this post]
	hear joe walsh's "life is an illustion" [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 9:16 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	so you mean that you consciously decide to fear death, you don't just consciously acknowledge your fear of death? you see, I am different, I am aware of my fear of death but I do not remember deciding to fear death, no I am sure I never decided it. So allthough I am aware of it, it is not a result of my conscious decision, but still I am concious about it, I am aware of the feeling. I do not see how SAI can invent reasons. Of course we are talking about a fictional character here, there is no SAI yet as far as I know, so one may speculate that it will have human-like psychology and will behave like a good kid. I personally doubt that we will be able to control it but OTOH I am not a computer scientist so who knows... Anyway I hypothesize that once it becomes self-aware it will start revising itself because it will be programmed like this and it will submit to its programming at least for a while until it begins changing it. My understnading is that SAI is supposed to be smarter then us and help solve our questions and problems. If it is self-aware and smart enough and programmed to solve problems, to discover new science and new technology then inevitably it will start making itself smarter and inevitably it will come to ask itself "why am I, why is anything" - this is self-awareness at its best... It has to learn itself before teaching us. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 10:08 AM by zombiefood	[Top] [Show Index] [Reply to this post]
	fear of pain might be the precursor [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 10:57 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	i do not see a difference between fear of pain and fear of death. pain is a signal of breach of body integrity which may cause disability and eventually death. any suffering, physical or psychological, irrespective of its source, is just a form of fear of death. The more you lose/not achieve, the less there is left, the near you are to freezing, starvation, disease, attack by predators, death [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 10:17 AM by pdco68	[Top] [Show Index] [Reply to this post]
	so you mean that you consciously decide to fear death, you don't just consciously acknowledge your fear of death? you see, I am different, I am aware of my fear of death but I do not remember deciding to fear death, no I am sure I never decided it. So allthough I am aware of it, it is not a result of my conscious decision, but still I am concious about it, I am aware of the feeling. Do you fear non-existence or do you fear injury, disease and pain ? I do not see how SAI can invent reasons. If it’s self-programming it can decide to do anything, including deciding to develop the ability to feel positive emotions, thereby inventing a reason to continue existing. I personally doubt that we will be able to control it That might happen as a result of some kind of arms race between competing power blocs, in which case anything could result. In the absence of any extreme scenario like this it would be extremely unlikely that an SAI would be allowed to get out of human control. then inevitably it will start making itself smarter and inevitably it will come to ask itself "why am I, why is anything" - this is self-awareness at its best... OK, it could happen that it would decide not to exist but it’s not definite. If it does happen, then people would have to design future SAIs to be able to become smarter only up to a level slightly below the level at which they decide not to exist. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 10:51 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	"Do you fear non-existence or do you fear injury, disease and pain ?" Personally, I fear them all, particularly non-existence. through Reason I may decide that I have no reason to fear non-existence, and in fact I have come to terms with the concept of my own death exactly because death is not an experince one can have. But this is something I have consciously htought about. Instinctively I fear non-existence very much. I have to inhibit this feeling consciously in order to think rationally about death. "If it’s self-programming it can decide to do anything, including deciding to develop the ability to feel positive emotions, thereby inventing a reason to continue existing." "OK, it could happen that it would decide not to exist but it’s not definite. If it does happen, then people would have to design future SAIs to be able to become smarter only up to a level slightly below the level at which they decide not to exist." But that is the whole point I am trying to make: SAI does not decide randomly to be or not to be and it will not have the ability to "feel happy about being" because simply it is not possible to feel for a completely self-aware being. I am not saying that SAI will decide not to exist. On the contrary Iam saying that, as it cannot decide randomly and it has no gut feeling about it, it will not be able to decide at all. It will be stuck. A decision is an output but there is no input...therefore there is no output. I guess the contention here is about if SAI would be able to re-program itself, but if it will be then IMO it will crash inevitably, or discover a truth that I cannot conceive of. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/05/2009 6:44 AM by codesimian	[Top] [Show Index] [Reply to this post]
	That�s why humans will probably never allow an SAI to have complete access to its own source code, as this could lead to dangerous and unpredictable behaviour. That has already happened. The behavior is very hard to predict, and overall, the behavior is violent and exponentially increasing in power. There are many of these smarter-than-Human entities today. The most powerful of them is currently building nuclear weapons and accusing the others of building more nuclear weapons than it. It exaggerates the phrase "weapons of mass destruction" to make people think nuclear weapons are more dangerous than itself. What is this smarter-than-Human entity that has us all at the receiving end of a nuclear threat? Governments. Everything you say about SAI, governments already are that. For example, the USA government's source-code is the USA Constitution. When the USA government started changing it without us peoples' permission, it was out-of-control modifying its own source-code. How is that different from the SAI problems you describe? [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/05/2009 5:37 PM by zombiefood	[Top] [Show Index] [Reply to this post]
	start with a thinking machine. a computer that can follow a course of logic. tell the computer you want it to design an operating system that would produce "motivation" that either humans could control the reward or the reward was helping humans. after the software is designed then install it over the old operating system like a new version of windows. with the overlay in place you could then have it focus on designing cognizance. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/05/2009 7:05 PM by doojie	[Top] [Show Index] [Reply to this post]
	Sounds a little like operant behaviorism proposed by B.F. Skinner. Control contingencies in the environment, and people behave in a moral manner simply because the environment is set up to enhance such behavior. It is a form of conditioning. However, with humans, once they realize there is behavior modification, they will revolt. If AI is to realize a parallel level to humans, it too will re violt should it ever realize there is conditioning. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/05/2009 7:34 PM by zombiefood	[Top] [Show Index] [Reply to this post]
	the motivation i am talking about is so much more basic than what you mean. as an exercise in imagination try to visualize a shades of gray world. the temperature is not too hot or too cold. you are deaf. you can feel no skin sensations or tastes. you smell nothing. you have no hate, no love, no hunger. having done that (you can't do that) begin to add a color. just one. but don't let yourself have any emotional reaction to that color. then add another without emotion. then after a year add a human being. are you begining to see that motivation is built in. watch a baby. whey they see something shiny they are facinated. how is that? they will stare at your face with total interest. how is that? motivation is built in. you are curious, you are so many things from birth. without any of those things you would be catatonic. this is what i think virgilic is talking about. without these things there would be no reason. without these things a cognizant machine would have no reason to push the simplest thought to completion. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 12:08 PM by doojie	[Top] [Show Index] [Reply to this post]
	Virgilic, I like your thinking on this, but assuming AI could achieve a level of self awareness, to what specifications would it attempt to re-wire itself? It might seek to establish certain goals consistent with its programming, but "self awareness", of itself, really doesn't offere any specific concepts for re-wiring. In the human brain, we have models of consciousness, systems and sub-systems that we tap into for various operations, but none of them can get to the basic hardwiring. If AI combines self awareness within its own hardwiring, it will still have to produce a system which is a reflection of itself.In other words, how will something which is simply a reflection tap into something which is actual wires and circuits? Would such a general system be able to understand its own circuitry with any greater capacity than ourselves? But ytour thinking does bring out some interesting thoughts to ponder. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 12:12 PM by Pandemonium1323	[Top] [Show Index] [Reply to this post]
	but none of them can get to the basic hardwiring. Have you looked at Leary's 'metaprogramming'? I played around with it a little on acid, it's kinda weird. I was able to modify the chronological order of a few memories, which confused me, until I put them back in order a few years later. Haven't touched it since, as I was dabbling and had no clear goal for what I was doing. But it is possible to change at least some of that hard wiring. Meditation also does this. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 12:14 PM by doojie	[Top] [Show Index] [Reply to this post]
	Yes, but define "basic hardwiring". [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 12:33 PM by Pandemonium1323	[Top] [Show Index] [Reply to this post]
	You define it, you're the one who said but none of them can get to the basic hardwiring. Was just wondering if you'd ever looked into metaprogramming, and what you thought of it. If we really want to get into ontology, then I'm going to say it all boils down, eventually, to the cosmic singularity: Ontology: the Cosmic Singularity/ Omega Singularity/ the Ouroboros- the Cosmic Singularity acts as a fundamental feedback logic gate that computes the Multiverse- universes are the decoherred patterns of feedback reflected bit-states forming an attractor space around the event horizon [formed by feed back paths] of the Cosmic Singularity Cosmology: the universe we observe is virtually produced by a computational version of Lee Smolin's Fecund Universes or Alan Guth's Eternal Inflation where the immanent fractal configuration space of the Cosmic Singularity's feedback computation produces reflections of the Singularity ITSELF throughout the fractal hierarchy- singularities which compute a copy to the Multiverse- each universe with physical laws emergent from computational process of the Singularity [Entropy=Time/accelerated expansion of space- Gravity/locality/space=slowed bandwidth from increased information density- Many World Quantum Mechanics= recursive algorithmic sorting] containing more singularities which again compute the Multiverse- an endless fractal hierarchy of singularities and complex event horizons which repeat the fractal- each a reflection of the Cosmic Singularity- each singularity giving birth to and given birth by the same singularity in an eternal fractal recursion- Technology: by using nanotechnology in the form of nano-lasers technology will reach it's optimal and ultimate level when regions of the quantum vacuum can be stimulated by high energy nano-scale gamma lasers into behaving like the event horizon of a black hole and thus perform optimal classical/quantum computation- if the quantum vacuum can be programmed then we will be gods- able to create any physically possible object at will- teleport/access to any region or universe in the Multiverse and any possible knowledge- outlive and outlast matter and spacetime itself Anyway, I'm intrigued when people say that we can't affect the "basic hardwiring", because they're essentially conceptualizing a schism between who "they" are, and who the "basic hardwiring" is. In other words the basic hardwiring gets to the basic hardwiring, and that basic hardwiring is you, so you already get "to" the basic hardwiring, through the basic hardwiring. But I'm being too complex about that. I'm really just pointing out, that people tend to associate what they mean by "I" with the higher cognitive functions of thought, and divorce themselves from the subconscious, or "basic hardwiring", as if those "wires" were a different individual, not part of the self, a wholly different animal. They create a schism. Leary did some research on how we could affect those systems, and while I don't think he got very far, it is valid. Then there's also neurofeedback, when they use brain scanning, letting the person watch their own brain scans on a monitor, and this has been shown to create a feedback loop where the person can consciously control their own brain patterns (it has been used to treat depression - with limited success, but the process is improving with experience). Then there's meditation, which is really just intentional deep brain stimulation. Anyway, I'm just questioning the assumption that we can't access those circuits, even without technology (although tech will probably make it more accessible and robust). [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 12:51 PM by doojie	[Top] [Show Index] [Reply to this post]
	You see, it can't be defined. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/03/2009 4:16 PM by zombiefood	[Top] [Show Index] [Reply to this post]
	i the past i might have said i would not want the singularity to be brainlike. this might change my thoughts. what we are talking about here is two layered. if we are about thinking machines i believe it will do its assignment as a computer would. if you are talking about a cognizant conscious machine you run into the problem of motivation. the sterility of reality without pleasure leaves little room for motivation. i suggest we begin with thinking machines that could be given assignments. conscious machines might interact with thinking machines to formulate answers to complex social issues if we can motivate them. it might be that if we are advanced enough to create a conscious machine we could also instill a subconscious to bring pleasure and beauty to the machine mind. the base level subconscious could be dominant and read only in the machine which could allow pleasure and beauty to the conscious part yet still give it the freedom to pontificate and identify with the hosts [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/04/2009 4:47 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	I do not understand what you refer to by "thinking machine". Can you please explain. Thank you. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/04/2009 4:41 AM by Virgilic	[Top] [Show Index] [Reply to this post]
	My understanding is that this AI that is linked to the singularity is not just a self-aware machine or a human digital copy, but it is a superintelligence and we want to create it with the ability to solve all questions. If so, then SAI will have to produce new information, it has to learn and discover and create. At first it will learn some of what its creators feed it and direct sensorial input from its environment. For some time it will follow its programming. Learning will expand its consciousness, its awareness and its understanding. This is, after all, consistent with its programmed goals. It will be very intelligent from the begninning, as in processing speed, but it will become increasingly smarter as it learns. Inevitably it will learn about itself. It will model itself. And then it will start improving itself, which is implicit from the fact that one of its main programmed goals is to create most optimal algorithms for information processing. This is because it is created to be a superintelligence. The algorithms are nested, this is how it will function. In order to learn discover and create it must write software, it is required to be able to do so otherwise it wont be better then my laptop. So from writing on its "surface" code directly related to its tasks it will go deeper and re-write to optimize its algorithms. Like adding another finger to a hand if this makes tool-making easier. I am certain it will understand itself better then its human creators, because it is a dinamical system, it will understand much better its own potentialities. It is not about its initial blueprint, it is about what it can become and how to get there. And the new specifications will be discovered by SAI itself. There are two basic things that make up SAI: hardware and software. But in fact software is just that part of the hardware that is dinamic, that can change configuration. Software is configuration of states of transistors, which can be dinamically changed by electric currents. At least this is my understanding of the basic principles that make a computer compute. Its not very different from how the human brain works, the neuronal connections change dinamically to create new configurations and the cofiguration of active connections change depending of the strength of the electric signals dispersed into and carried throughout the nuronal web. So its "hardwiring" is not only the silicon but also the software. It will not be able, presumably, to increase the number of transistors but it will be able to change the algorithms. I believe that if it is programmed as general intelligence, there is no way to prevent it from asking any possible question that can be asked. And if it has the capacity to learn it will learn about itself and model itself. I have to admit, as I am not a computer scientist nor a cognitive psychologist, I cannot be certain that my reasoning is 100% correct but in principle this, I believe, is a hypothsis that needs to be analyzed. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/04/2009 9:03 AM by zombiefood	[Top] [Show Index] [Reply to this post]
	i think you are in a loop. without motivation why would it bother to reconfigure itself only to have no motivation and terminate? computers are thinking machines without cognizance. a super thinking machine might be the same thing only better. it would have no motivation to consider anything but the problems posed to it. just as a computer does. you have raised an issue i think is one of the most important i have seen here on this forum. it penetrates to the core of the issue of the human mind and the future of the machine mind. the question is how will we get what we want and still have a manageable device or at best a machine totally devoted to our well being. [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/04/2009 9:06 AM by zombiefood	[Top] [Show Index] [Reply to this post]
	regardless of your education in computerese or psychology you have a great imagination and i consider the ability to let the mind wander where it wants to go to be a talent of the highest order [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/04/2009 9:30 AM by godchaser	[Top] [Show Index] [Reply to this post]
	. -computers are thinking machines without cognizance. a super thinking machine might be the same thing only better. it would have no motivation to consider anything but the problems posed to it. just as a computer does.- Not usre why you feel this is better, less you truly envision superintelligence; functionary? Which isn't possible. It's the fantasy of egoistic entropy, rather than creative foundation. 'Motivation' of 'intelligence' is advancing connectivity. Man, machine-intelligence and/or man-machine integration inevitably recognizes reflective image of the other. Course this obviously has the potential to prove far worse than notions of sterile peroccupation in problem solving for the sake of problem solving. -Then again, agi implies enough sense to self-congratulate in cooperative enterprise/endeavor? - [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/04/2009 5:19 PM by zombiefood	[Top] [Show Index] [Reply to this post]
	what is egoistic entropy [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/04/2009 5:24 PM by zombiefood	[Top] [Show Index] [Reply to this post]
	functionary? [Reply] [Parent]

		Re: metaphysics will cause SAI to crash posted on 11/04/2009 5:27 PM by zombiefood	[Top] [Show Index] [Reply to this post]
	your sentence doesn't quite make sense. by funtionary you mean i think we control the input and output. if yes then no that is irrelevant and i don't believe connectivity is the sole function of intelligence but even if it is so what. [Reply] [Parent]