Hi Ben Hereby my proposed additional topics / references for your wiki - aimed at the more computer scienty/mathematically challenged (like me): Sorry don't have the time to add directly to the wiki
AGI ARCHITECTURES (EXPANDS on the COGNITIVE ARCHITECTURES section) Questions about any Would-Be AGI System. Ben Goertzel - May 20, 2002 Artificial General Intelligence: A Gentle Introduction - Pei Wang Architectures for intelligent systems by J. F. Sowa Cognitive Architectures: Research Issues and Challenges by Langley, Laird & Rogers. Choosing and getting started with a cognitive architecture to test and use human-machine interfaces by Frank RITTER. MMI-Interaktiv, #7, Jun04 "Artificial General Intelligence" PowerPoint presentation by ?? "Artificial General Intelligence" - Goertzel, Ben; Pennachin, Cassio (Eds). Chapter by Peter Voss Four Contemporary AGI Designs: a Comparative Treatment Sep2006 Stan FRANKLIN, Ben GOERTZEL, Alexei SAMSONOVICH, Pei WANG Mixing Cognitive Science Concepts with Computer Science Algorithms and Data Structures: An Integrative Approach to Strong AI. Moshe Looks & Ben Goertzel Computational ARchitectures for Intelligence and Motivation Darryl N. Davis The 17th IEEE International Symposium on Intelligent Control , ISIC’02, Canada, Oct 2002 Considerations Regarding Human-Level Artificial Intelligence - Nils J. Nilsson - Jan 2002 A Survey of Artificial Cognitive Systems: Implications for the Autonomous Development of Mental Capabilities in Computational Agents David Vernon. AGENTS Search and select depending on the nature of your AGI architecture. AUTONOMIC COMPUTING Any one of the IBM AC overview papers e.g. Practical Autonomic Computing: Roadmap to Self Managing Technology A White Paper Prepared for IBM January 2006 BOTS Read any document on AIML, check out on the Loebner prize and check the source code of at least one ChatterBot in your preferred programming langague. COGNITION List of cognitive biases - Wikipedia Contemporary Approaches to Symbol Grounding - Moshe Looks Interior Grounding, Reflection, and Self-Consciousness - Marvin Minsky Intl Conf on Brain Mind & Society, Japan 2005. Solving the Symbol Grounding Problem: a Critical Review of Fifteen Years of Research by Mariarosaria Taddeo and Luciano Floridi French, R. M. (2002). The Computational Modeling of Analogy -Making. Trends in Cognitive Sciences, 6(5), 200-205. COMPLEXITY THEORY - any good overview COMPUTATIONAL INTELLIGENCE Feigenbaum - Grand Challenges for Computational Intelligence Some paper on AIXI Craenen & Eiben - Computational Intelligence 2002 Moshe Looks - Learning with Semantic Spaces: From Parameter Tuning to Discovery One of Wang's papers on Cognitive Informatics e.g. Theoretical Framework of CI, CI Models of the Brain or similar INTRODUCTORY, POPULAR & GENERAL BOOKS: # Eric Baum, What is Thought?, 2004 # Ben Goertzel, The Hidden Pattern: A Patternist Philosophy of Mind, 2006 # Ben Goertzel, Cassio Pennachin (Eds.), Artificial General Intelligence, 2007 # Jeff Hawkins, On Intelligence, 2004 # Steven Pinker, The Stuff of Thought, 2007; How the Mind Works; The language Instinct. # Storrs Hall, Beyond AI # Franklin, Artificial Minds # Sowa, Knowledge Representation # Douglas Hofstadter, Godel, Escher & Bach # Ray Kurzweil, The Age of Spritiual Machines & The Singularity Is Near # Wolfram, A new kind of science # Smith, On the origin of objects # Jeff Hawkins, on intelligence and at least two of the following 'cognitive science compilations' # Rosenthal (ed), The Nature of Mind # Bechtel & Graham, a Companion to Cognitive Science # Posner (ed), Foundationsof Congitive Science # Wilson & Kehl, THe MIT Encyclopedia of Cognitive Sciences Some other popular science titles that you might consider (some are quite dated): #Kevin Warwick, In the mind of the Machine #Robert Winston, the human mind #Philip Johnson-Laird - the computer and the mind #Gardenfors - conceptual spaces #Rita carter - mapping the mind #Rondey Brooks - Robot #and you should read at least one book by Roger Penrose (and/or perhaps Daniel Dennett.). Some advice on literature/articles NOT to read i.e. TIME-CONSUMING debates to avoid wasting your precious time on: ** PROGRAMMING LANGUAGES ** If you're defining your own AGI project, you have to choose (ideally one) language. IMHO if you already have lots of experience and feel comfortable in a particular language and it *seems to you* that it is adequate, then don't waste time debating other languages - *all* languages have their advantages and limitations. However, if you have to choose a new language (or don't mind changing) then: if you need raw speed and current hardware is likely to be a bottleneck, then: = if your algorithms are fairly classic in nature, choose C#, C++ or similar - ideally a dialect that supports parallel hardware architectures = but if your algorithms are fairly esoteric and/or you have 'strange' data structures, looks at Lisp, SmallTalk or similar if you think hardware speed is not an issue or you're prototyping or you're not such as great programmer, choose a scripting language such as Python (I am a great fan of its power, libraries and its great productivity due to conciseness) More likely you'll be cooperating with an existing team then it's betst to fit in with (one of) their language(s). However, you MUST have a good set of reference materials to YOUR programming language e.g. for me/Python: Core Python Programming (Chun), Python Essential Reference (Beazly), Python Cookbook (Martelli et al), Python Standard Library (Lundb), Text Processing with Python (Lutz), and the Introduction to Natural Language Programming (really an NLTK & Python manual, by Bird et al) along with lots of articles on finer programming issues (eg introspection, metaprogramming, minilanguages, etc.) ** PHILOSOPHICAL DEBATES ** Do not I repeat most definitely NOT get sucked in by the vast amount of philosophical literature by philosophers like Searle who claim AGI is impossible. I am not saying that the debate is not interesting or doesnt have its merits or that you have to be unaware of it. However, it is like a faith, you either are on the one side or on the other. If you still need to be convinced that AGI is possible, then you're in the wrong field. No philosophical argument is going to convince you either way - it's a position you intuit (or not, as the case may be). Pretty much similar to trying to convince someone to change religion by means of theological debate. Not that it doesn't make for interesting reading or mind exercise, but most of it is a matter of semantics. Having said that, a number of philosophers are liely to shed some light on someof the issues you may be thinking about e.g. I like many of Carruther's articles on the modularity of mind. ** SYMBOLIC vs CONNECTIONISM vs STATISTICAL vs ... DEBATE ** It is essential to be acutely aware and knowledgeable of the different paradigms in AI and their major achievements. An excellent overview book in this regard is BODEN's historical account of AI. However do not enter the *debate* about whether symbolic, connectionist, statistical, genetic algorithm etc. is the 'one and only' best one. It depends on your architecture, approach and problem domain. Trying to read too much architecture that blindly advocates one approach only is pretty useless IMHO. Once you are aware of the algorithms, capabilities and achievements of each approach decide on those that will work for you and investigate those more thoroughly. Finally be selective on whom you engage with on the AGI list ;-) ------------------------------------------- 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=8660244&id_secret=98558129-0bdb63 Powered by Listbox: http://www.listbox.com
