RE: Re[8]: [agi] Funding AGI research
From: Dennis Gorelik [mailto:[EMAIL PROTECTED] John, Is building the compiler more complex than building any application it can build? Note, that compiler doesn't build application. Programmer does (using compiler as a tool). Very true. So then, is the programmer + compiler more complex that the AGI ever will be? Or at some point does the AGI build and improve itself. John - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=70653218-951955
Re: Re[8]: [agi] Funding AGI research
I don't think we yet know enough about how DNA works to be able to call it a conglomerated mess, but you're probably right that the same principle applies to any information system adapting over time. Similarly the thinking of teenagers or young adults is sometimes quite clear (almost cartoon-like) but as they get older all sorts of exceptions and contradictions creep into the thought process. On 28/11/2007, John G. Rose [EMAIL PROTECTED] wrote: A stretch of an analogy is human DNA. Most of it is a conglomerated mess (90%+?) but somehow the grand human design result comes out of it. - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=69340140-45577e
RE: Re[8]: [agi] Funding AGI research
From: Bob Mottram [mailto:[EMAIL PROTECTED] I don't think we yet know enough about how DNA works to be able to call it a conglomerated mess, but you're probably right that the same principle applies to any information system adapting over time. Similarly the thinking of teenagers or young adults is sometimes quite clear (almost cartoon-like) but as they get older all sorts of exceptions and contradictions creep into the thought process. It does happen too with academia where there is this nice picture of how things should work but then reality is different. Software is just weird and has unpredictable qualities different from other forms of engineering. There are situations with software where money is just thrown at it lavishly over and over defying any sort of reasonableness, example the VC's friends son has this great idea, they call the software GaGa (they make it sound like Google on purpose, happens all the time) and they throw money at it and sell the company and the software winds up doing something totally different from what was originally planned or sometimes it just becomes vaporware. Since much software is in many ways non-material and mutateable it is treated thus. Internally used and developed software within companies, many times the software that runs the companies, can take extremely bizarre twists of fate... You can say AGI software is special, and it is. If its purpose and goals can be maintained enough, like in specialized software such as weather modeling software, it can stay on course. Yet AGI is very associated with narrow AI so the likelihood of business needs interrupts occurring is high. Also humans are building it and we have special needs that take priority ofttimes. John - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=69497333-d9b7d1
Re: Re[8]: [agi] Funding AGI research
My claim is that it's possible [and necessary] to split massive amount of work that has to be done for AGI into smaller narrow AI chunks in such a way that every narrow AI chunk has it's own business meaning and can pay for itself. You have not addressed my claim, which has massive evidence in the history of AI research to date, that narrow AI chunks with AGI compatibility are generally much harder to build than narrow AI chunks intended purely for standalone performance, and hence will very rarely be the best economic choice if one's goal is to make a narrow-AI chunk serving some practical application within (the usual) tight time and cost constraints. -- Ben G - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=69291210-f650cd
RE: Re[8]: [agi] Funding AGI research
My claim is that it's possible [and necessary] to split massive amount of work that has to be done for AGI into smaller narrow AI chunks in such a way that every narrow AI chunk has it's own business meaning and can pay for itself. You have not addressed my claim, which has massive evidence in the history of AI research to date, that narrow AI chunks with AGI compatibility are generally much harder to build than narrow AI chunks intended purely for standalone performance, and hence will very rarely be the best economic choice if one's goal is to make a narrow-AI chunk serving some practical application within (the usual) tight time and cost constraints. I'd like to comment here - this all makes sense BUT - software for many reasons, and after myself watching and working with software of many types for many years, I'm still trying to figure it out - defies logic. Especially internally developed applications within companies, those applications which survive over the years, don't follow behaviors that you'd think they should. Many are conglomerated messes. The crystal clear software dies often. Why? Because it has to adapt to extreme conditions - basically business demands and interjections of business needs overpowers the forces of perfect engineering. And there are other forces, human, material, etc.. That's one of the reasons why I would say that somebody that creates an AGI design and gets x million $ startup cap will have a higher chance of failing than a software that has to slug it out to survive by becoming a conglomerated narrow AI/AGI hybrid mess. It's just weird and this happens often but not all the time of course. A stretch of an analogy is human DNA. Most of it is a conglomerated mess (90%+?) but somehow the grand human design result comes out of it. John - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=69303611-cb1ba7
Re: Re[8]: [agi] Funding AGI research
Hi, On 20/11/2007, Dennis Gorelik [EMAIL PROTECTED] wrote: Benjamin, That's massive amount of work, but most AGI research and development can be shared with narrow AI research and development. There is plenty overlap btw AGI and narrow AI but not as much as you suggest... That's only because that some narrow AI products are not there yet. Could you describe a piece of technology that simultaneously: - Is required for AGI. - Cannot be required part of any useful narrow AI. Oh please. That's like saying that the theory of areodynamics is the same for fast cars and for airplanes (it is), so lets build a fast car, and we'll probably have an airplane come out of it as a side-effect. If you really want to build something, focus on building that thing. As long as you focus on something else, you will fail to take the needed steps to get to the objective you want. To be more direct: a common example of narrow AI are cruise missles, or the darpa challange. We've put tens of millions into the darpa challange (which I applaud) but the result is maybe an inch down the road to AGI. Another narrow AI example is data mining, and by now, many of the Fortune 500 have invested at least tens, if not hundreds of millions of dollars into that .. yet we are hardly closer to AGI as a result (although this business does bring in billions for high-end expensive computers from Sun, HP and IBM, andd so does encourage one component needed for agi). But think about it ... billions are being spent on narrow AI today, and how did that help AGI, exactly? --linas - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=68370711-0257fe
Re: Re[8]: [agi] Funding AGI research
On 21/11/2007, Dennis Gorelik [EMAIL PROTECTED] wrote: Benjamin, That's massive amount of work, but most AGI research and development can be shared with narrow AI research and development. There is plenty overlap btw AGI and narrow AI but not as much as you suggest... That's only because that some narrow AI products are not there yet. Could you describe a piece of technology that simultaneously: - Is required for AGI. - Cannot be required part of any useful narrow AI. My theory of intelligence is something like this. Intelligence requires the changing of programmatic-structures in an arbitrary fashion, so that we can learn, and learn how to learn. This is because I see intelligence as the means to solve the problem solving problem. It does not solve one problem but changes and reconfigures itself to solve whatever problems it faces, within its limited hardware/software and energy constraints. This arbitrary change can result in the equivalent of bugs and viruses, this means there needs to be ways for these to be removed and prevented from spreading. This requires there be a way to distinguish good programs from bad, so that the good programs are allowed to remove bugs from others, and the bad programs prevented from being able to alter other programs. Solving this problem is non-trivial and requires thinking about computer systems in a different way to other weak AI problems. Narrow AI is generally solving a single problem, and so does not need to change so drastically and so does not need the safeguards. It can just concentrate on solving its problem. Will Pearson - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=67564879-97ae32
Re: Re[8]: [agi] Funding AGI research
William P: My theory of intelligence is something like this. Intelligence requires the changing of programmatic-structures in an arbitrary fashion, so that we can learn, and learn how to learn. Well, you're getting v. close. But be careful, because you'll upset Ben and Pei not to mention cog sci. The moment you make a mechanical mind arbitrary to any extent, it ceases to be deterministic. Tch tch. And the moment you make the application of programs arbitrary, well, they cease to be programs in any true sense. Shock, horror. Perhaps the only way such a mind could function is if it only had a rough idea rather than a precise set of programmed instructions for how to get from A to Z and conduct any activity - a precis rather than a program of what to do - and would have to freely/arbitrarily combine steps and sub-routes to see/learn what worked and reached the goal. As scientists do. And technologists do. And computer programmers in writing their programs do. And human beings do period. Yes, that would require intelligence in the full sense. P.S. And, as you indicate, such a machine would only have a rough idea of how to *learn* as well as directly conduct an activity - it wouldn't have any preprogrammed set of instructions for learning and correcting mistakes, either. 'What then, I thought myself, if I [Robot Daneel Olivaw] were utterly without laws as humans are? What if I could make no clear decision as to what response to make to some given set of conditions? It would be unbearable, and I do not willingly think of it. Isaac Asimov, Robots and Empire - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=67599979-79e74d
Re: Re[8]: [agi] Funding AGI research
--- Dennis Gorelik [EMAIL PROTECTED] wrote: Could you describe a piece of technology that simultaneously: - Is required for AGI. - Cannot be required part of any useful narrow AI. A one million CPU cluster. -- Matt Mahoney, [EMAIL PROTECTED] - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=67809185-13d25e
Re: Re[8]: [agi] Funding AGI research
On Nov 22, 2007 12:59 AM, Matt Mahoney [EMAIL PROTECTED] wrote: --- Dennis Gorelik [EMAIL PROTECTED] wrote: Could you describe a piece of technology that simultaneously: - Is required for AGI. - Cannot be required part of any useful narrow AI. A one million CPU cluster. Is a required part of Google, which is very useful narrow AI. The main piece of technology I reckon is required to make more general progress is a software framework, which would be useful for narrow AI but is only essential if you want to go beyond that. - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=67839995-f106a4
Re: Re[8]: [agi] Funding AGI research
Could you describe a piece of technology that simultaneously: - Is required for AGI. - Cannot be required part of any useful narrow AI. The key to your statement is the word required Nearly any AGI component can be used within a narrow AI, but, the problem is, it's usually a bunch easier to make narrow AI's using components that don't have any AGI value... Another way to go -- use existing narrow AIs as prototypes when building AGI. I don't really accept any narrow-AI as a prototype for an AGI. This is an example of what I meant when I said that what counts as a prototype is theory-dependent, I suppose... I think there is loads of evidence that narrow-AI prowess does not imply AGI prowess, so that a narrow-AI can't be considered a prototype for an AGI.. -- Ben G - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=67477308-6a7310