Re: [agi] Re: Merging - or: Multiplicity
2008/5/27 Mike Tintner [EMAIL PROTECTED]: Will:And you are part of the problem insisting that an AGI should be tested by its ability to learn on its own and not get instruction/help from other agents be they human or other artificial intelligences. I insist[ed] that an AGI should be tested on its ability to solve some *problems* on its own - cross-domain problems - just as we do. Of course, it should learn from others, and get help on other problems, as we do too. But you don't test for that, and as the loebner prize shows you only tend to get what you test for. But if it can't solve many general problems on its own - which seemed OK by you (after setting up your initially appealing submersible problem - solutio interrupta!) - then it's only a narrow AI. I am happy for the baby machine (which is what we will be dealing with to start with) not to be able to solve general problems on its own. Later on I would be disappointed. Will Pearson --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Design Phase Announce - VRRM project
2008/5/27 Steve Richfield [EMAIL PROTECTED]: William, This sounds like you should be announcing the analysis phase! Detailed comments follow... Design/research/analysis, call it what you will. On 5/26/08, William Pearson [EMAIL PROTECTED] wrote: VRRM - Virtual Reinforcement Resource Managing Machine Overview This is a virtual machine designed to allow non-catastrophic unconstrained experimentation of programs in a system as close to the hardware as possible. There have been some interesting real machines in the past, e.g. the Burroughs 5000 and 6000 series computers that seldom crashed. When they did, it was presumed to be either an OS or a hardware problem. At Remote Time-Sharing we extended these in a virtual machine, to make a commercial time-sharing system that NEVER EVER crashed after initial debugging. This while servicing secondary schools in the Seattle Area with many hackers, including a very young Bill Gates and Paul Allen. Systems now crash NOT because of the lack of some whiz-bang technology, but because architectural development has been in a state of arrested development for the last ~35 years. It is not just crashes that I worry about but memory corruption and other forms of subversion. This should allow the system to change as much as is possible and needed for the application under consideration. Currently the project expects to go to the operating system level (including experimentation on schedulers and device drivers). A separate sub-system supplies information on how well the experiment is going. The information is made affective by making it a form of credit periodically used to bid for computational system resources and to pass around between programs. This sounds like a problem for real-time applications. In what sense? Expected deployment scenarios - Research and possible small scale applications on the following - Autonomous Self-managing robotics - A Smart operating system that customises itself to the users preferences without extensive knowledge on the users part Language - C Whoops, there are SERIOUS limitations to what can be made reliable in C. C is purely the language for the VRRM, what the programs will be implemented inside the VM is completely up to the people that implement them. Progress Currently I am hacking/designing my own, but I am open to going to a standard machine emulator if that seems easy at any point. I expect to heavily re-factor. I am focussing on the architectural registers, memory space and memory protection first and will get on to the actual instruction set last. This effort would most usefully be merged with the 10K architectures that I have discussed on this forum. Merging disparate concerns might actually result in a design that someone actually constructs. Possibly after I have completed the VRRM and tested it to see if it works how I think it works. But silicon implementation is not on the agenda at the moment. I'm also in parallel trying to design a high level language for this architecture so the internals initial programs can be cross-compiled for it more easily. Does this require a new language, or just some cleverly-named subroutines? A different set of system calls in the least. Some indication of how important the memory is in dynamic memory creation is needed for example. Current Feature plans - Differentiation between transient and long term storage to avoid unwanted disk thrashing Based on the obsolete concept of virtual memory rather than limitless RAM. We don't have limitless RAM, and I won't be implementing virtual memory. snip because I don't have time - Specialised Capability registers as well as floating point and integers Have you seen my/our proposed improvements to IEEE-754 floating point, that itself incorporates a capability register?! Perhaps we should look at a common design? Do you mean capability in the same sense as me? http://en.wikipedia.org/wiki/Capability-based_security Will Pearson --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Re: Merging - or: Multiplicity
Steve: I have been advocating fixing the brain shorts that lead to problems, rather than jerking the entire world around to make brain shorted people happy. Which brain shorts? IMO the brain's capacity for shorts in one situation is almost always a capacity for short-cuts in another - and dangerous to tamper with. Steve:Let's instead 1. make something USEFUL, like knowledge management programs that do things that people (and future AGIs) are fundamentally poor at doing Well, in principle, a general expert system that can be a problem-solving aid in many domains would be a fine thing. But - if you'll forgive the ignorance of this question - my impression was that expert systems were a big fad that has largely failed??? If you have a link to some survey here, I'd appreciate it. Steve, the capacity for general thinking/intelligence HAS to be - and is being - explored. William may be right that all the main AGI-ers are like him avoiding the challenge of general problemsolving, and hoping that the answer will emerge later on in the development of their systems. But roboticists are setting themselves general problems nbw - in the shape if nothing else of the ICRA challenge, as I've pointed out before. I --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: Merging threads was Re: Code generation was Re: [agi] More Info Please
Steve/Stephen: I am planning to archive all conversations .This is pretty simple with text, but when things move into real-time moving images from which to understand the world, this takes a little more storage. No one's yet actually trying to develop movie AI/AGI - an intelligence that can live in and/or respond to a continuous movie[s] of the world, are they? Ben's system, from the v. little I saw, gestures at this, but falls short. --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: Merging threads was Re: Code generation was Re: [agi] More Info Please
2008/5/28 Mike Tintner [EMAIL PROTECTED]: No one's yet actually trying to develop movie AI/AGI - an intelligence that can live in and/or respond to a continuous movie[s] of the world, are they? Ben's system, from the v. little I saw, gestures at this, but falls short. I'm doing stuff with robotics which is mostly about processing sequences of images (I call the offline playbacks used for parameter optimisation dream sequences), although probably what I'm doing doesn't qualify as AGI in a strict sense - it's more reminiscent of the Grand/Urban Challenge stuff. --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: Merging threads was Re: Code generation was Re: [agi] More Info Please
Bob: I'm doing stuff with robotics which is mostly about processing sequences of images (I call the offline playbacks used for parameter optimisation dream sequences), although probably what I'm doing doesn't qualify as AGI in a strict sense - it's more reminiscent of the Grand/Urban Challenge stuff. Sounds interesting. Can you give us a little more detail (or link). What kind of robot, where? Doing what? Watching what movie? And how does it dream - optimise/correct actions? --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: Merging threads was Re: Code generation was Re: [agi] More Info Please
2008/5/28 Mike Tintner [EMAIL PROTECTED]: Sounds interesting. Can you give us a little more detail (or link). What kind of robot, where? Doing what? Watching what movie? And how does it dream - optimise/correct actions? Link: http://code.google.com/p/sentience/ A picture of the robot: http://farm4.static.flickr.com/3280/2427779514_d28b368557_b.jpg --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Design Phase Announce - VRRM project
William, On 5/27/08, William Pearson [EMAIL PROTECTED] wrote: 2008/5/27 Steve Richfield [EMAIL PROTECTED]: Systems now crash NOT because of the lack of some whiz-bang technology, but because architectural development has been in a state of arrested development for the last ~35 years. It is not just crashes that I worry about but memory corruption and other forms of subversion. This all is the result of poor design. For example, the old Burroughs hardware performed all computed-address operations via array descriptors, which not only showed where the array began, but also what all of its dimensions are. Where multiple subscripts were used, the multiplication needed to be done to compute the address was all done in the hardware by implication, and only after each of the bounds were checked. Note that this costs NO more time when the additional hardware needed to perform this in parallel is included in the design. This should allow the system to change as much as is possible and needed for the application under consideration. Currently the project expects to go to the operating system level (including experimentation on schedulers and device drivers). A separate sub-system supplies information on how well the experiment is going. The information is made affective by making it a form of credit periodically used to bid for computational system resources and to pass around between programs. This sounds like a problem for real-time applications. In what sense? Because real-time applications typically need whatever they need RIGHT NOW. Windows is woefully inadequate for real-time applications, as I discovered when I interfaced Dragon NaturallySpeaking to my AI program. It would often stop for many seconds while new things were paged into memory, run slow as complex components were kicking each other out of cache, etc. Whoops, there are SERIOUS limitations to what can be made reliable in C. C is purely the language for the VRRM, what the programs will be implemented inside the VM is completely up to the people that implement them. Progress Currently I am hacking/designing my own, but I am open to going to a standard machine emulator if that seems easy at any point. I expect to heavily re-factor. I am focussing on the architectural registers, memory space and memory protection first and will get on to the actual instruction set last. This effort would most usefully be merged with the 10K architectures that I have discussed on this forum. Merging disparate concerns might actually result in a design that someone actually constructs. Possibly after I have completed the VRRM and tested it to see if it works how I think it works. But silicon implementation is not on the agenda at the moment. My whole point was that reasonable architectures would eliminate some/many of the problems that you seek to fix. I'm also in parallel trying to design a high level language for this architecture so the internals initial programs can be cross-compiled for it more easily. Does this require a new language, or just some cleverly-named subroutines? A different set of system calls in the least. Some indication of how important the memory is in dynamic memory creation is needed for example. Isn't this just a small-RAM problem? Wouldn't a few more gigs of RAM obviate this, and be available before your software is completed? Current Feature plans - Differentiation between transient and long term storage to avoid unwanted disk thrashing Based on the obsolete concept of virtual memory rather than limitless RAM. We don't have limitless RAM, It's getting pretty close. and I won't be implementing virtual memory. snip because I don't have time - Specialised Capability registers as well as floating point and integers Have you seen my/our proposed improvements to IEEE-754 floating point, that itself incorporates a capability register?! Perhaps we should look at a common design? Do you mean capability in the same sense as me? Perhaps not, but shouldn't all capabilities be kept together? Bits would enable various features and differences, e.g. gradual overflow, logarithmic representation, significance representation, etc. This can be a global characteristic when a program is communicating floating-point to other programs. Steve Richfield --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
[agi] Adaptivity in Hybrid Cognitive Systems Osnabruck PhD program
I'm not affiliated but I've found this interesting. They seem to have 8 positions for PhD students: http://www.cogsci.uni-osnabrueck.de/PhD/GK/ Their research program is really worth checking-out: http://www.cogsci.uni-osnabrueck.de/PhD/GK/research/body.html --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
[agi] U.S. Plan for 'Thinking Machines' Repository
Fellow AGI-ers, At the risk of being labeled the list's newsboy... U.S. Plan for 'Thinking Machines' Repository Posted by samzenpus on Wednesday May 28, @07:19PM from the save-those-ideas-for-later dept. An anonymous reader writes Information scientists organized by the U.S.'s NIST say they will create a concept bank that programmers can use to build thinking machines that reason about complex problems at the frontiers of knowledge - from advanced manufacturing to biomedicine. The agreement by ontologists - experts in word meanings and in using appropriate words to build actionable machine commands - outlines the critical functions of the Open Ontology Repository (OOR). More on the summit that produced the agreement here. Cheers, Brad --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Consciousness vs. Intelligence
--- John G. Rose [EMAIL PROTECTED] wrote: Consciousness with minimal intelligence may be easier to build than general intelligence. General intelligence is the one that takes the resources. A general consciousness algorithm, one that creates a consciousness in any environment may be simpler that a general intelligence algorithm that acquires intelligence in any environment. The two can go hand in hand but one can be minimized against the other. But I don't understand the relationship between consciousness and intelligence. I want to say that they are like disjoint vectors but that doesn't seem right... You need to define your terms. What properties of an algorithm make it conscious? What properties make it intelligent? To some people, the two terms are equivalent. To others, consciousness does not exist. -- Matt Mahoney, [EMAIL PROTECTED] --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: Competitive message routing protocol (was Re: [agi] Deliberative vs Spatial intelligence)
--- J. Andrew Rogers [EMAIL PROTECTED] wrote: There is an increasingly strong political incentive (between countries) to create distributed indexes, but quite frankly the technology does not exist. This was something I studied in earnest when various governments started demanding such guarantees. To the best of my knowledge, we do not have mathematics that can support the guarantees desired, though decentralized indexes are certainly practical if one ignores certain considerations that are politically important. I would think that governments would be threatened by a distributed index. It would make file sharing networks useful to the point of replacing the client-server model. There would be no way to block undesirable material such as copyrighted movies, child porn, jihadist literature, etc. But that's not really my interest. I'm interested in making the internet more useful. I agree we don't have the mathematics to guarantee results. My arguments are based on economics. Peers have an economic incentive to cooperate, to supply useful and accurate information, and establish a reputation so they can sell advertising. They have an incentive to use the same language as most other peers, so the protocol should converge. They have an incentive to specialize in niches left unoccupied by others, so the collection becomes an AGI. Something to understand about the big server clusters: as commonly implemented, the online server cluster is independent of the content generation cluster. Queries may be very cheap to serve even if the aggregation and analytics process is expensive. Compute a result once and serve it to the world a thousand times. The real problems occur when the data set is not sufficiently static that this trick is plausible. Fortunately, no one has noticed the man behind the curtain (yet). That is the problem I am addressing. CMR makes no distinction between queries and documents. They are just messages. It does not matter which is posted first. If they are a close match, then each goes to the originator of the other. You could initiate an interactive conversation by posting a message about anything to nobody in particular, and it will go to anyone who cares. Losing to Google is predicated on following their path, and they occupy a space where the computer science is transparently inadequate. It does not take much of a qualitative shift in the market to kill a company in that position. There is plenty of vulnerability left in the market. I think Google will initially be a peer with a high reputation, but they will have to adapt to the new model. CMR needs to interact cleanly with the existing web in order for it to take off. A simple peer might just forward anything that looks like a query to Google and maybe a few other search engines. I would argue, from a business perspective, is that most of the value with respect to distribution is in the metadata protocol, virtually all of which are based on naive designs that ignore literature in practice. A really strong metadata protocol that could be standardized would generate a hell of a lot of value. Past that, whoever controls the essential data under that protocol would win, and for better or worse, Google is largely not responding to this. There are many types of data they have no capacity to handle in bulk. This is not so much a criticism of Google but an observation about their actual behavior. Nobody will control any critical part of the network. That is the beauty of it. -- Matt Mahoney, [EMAIL PROTECTED] --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com