We hope you are planning to join us for AAAI-08 and IAAI-08 in Chicago, Illinois, July 13-17, 2008. The AAAI-08 program will feature Eric Horvitz's cal papers will be highlighted as The program will include a research AAAI presidential address, as well as exceptional papers during the conference-wide track, industry track, invited speakers, five outstanding invited talks. Registration information invited speakers include Alexei A. Efros July 16, and another 23 short and other program details will (Carnegie Mellon University) whose papers will be presented as posters. Using Lots of Data to Infer Geometric, and awards will continue for its aiide08.php Please Photometric and Semantic Scene Properties second year with all the Hollywood send inquiries to aiide08@aaai.org

Humans learn from their mistakes. When things go badly, we notice that something is amiss, figure out what went wrong and why, and attempt to repair the problem. Artificial systems depend on their human designers to program in responses to every eventuality and therefore typically don’t even notice when things go wrong, following their programming over the proverbial, and in some cases literal, cliff. This article describes our past and current work on the Meta-Cognitive Loop, a domain-general approach to giving artificial systems the ability to notice, assess, and repair problems. The goal is to make artificial systems more robust and less dependent on their human designers.

Because the data consisted of long records of real values, the student was advised to use artificial neural networks. After several weeks of producing random classifiers, the student showed up at my office and asked whether I could help. It always seems a good idea to analyze the data first, so we constructed a primitive visualization: signal strength of four antennae over time. The graphs looked like we'd glued a pen on a dog's tail while showing him a juicy T-bone steak. I suggested we add a few functions, such as pairwise difference, mean, deviation, and so on--just to get a feel for the data.

There have been many erroneous pre-scientific and common sense inductions. We want to understand why people believe in wrong theories. Our hypothesis is that mistaken inductions are due not only to the lack of facts, but also to the poor description of existing facts and to implicit knowledge which is transmitted socially. This paper presents several experiments the aim of which is to validate this hypothesis by using machine learning and data mining techniques to simulate the way people build erroneous theories from observations.

Narrowly focused task and domain specific AI has been applied successfully for more than twenty five years, and has produced immense value in industry and government. It doesn’t lead directly to artificial general intelligence (AGI), but it does have real problem solving value. It is useful to note that many of the reasons why some otherwise meritorious AI applications fail have nothing to do with the AI per se, but rather, with systems engineering and organizational issues. For example: the domain expert is pulled out to work on more critical projects; the application champion rotates out of his/her position; or the sponsor changes priorities. A system may not make it past an initial pilot test for logistical vs. substantive technical reasons. Some embedded AI systems may work well for years on a software platform that is orphaned and porting it would be prohibitively expensive. A system may work well in a pilot test, but it might not scale for huge numbers of users without extensive performance optimization. The core AI system may be great but the user interface could be suboptimal. The delivered application system might work well, but it could be hard to maintain internally. The system may work according to the sponsor’s requirements, but it might not be applied to the part of the problem that delivers the largest economic results; or the system might not produce enough visible organizational benefits to protect it in subsequent budget battles. Alternatively, the documented results may be quite strong, but may not be communicated effectively across organizational boundaries. All software projects are vulnerable to one or more of these problems. The fact that some software projects have a relatively small percentage of their total code in embedded AI methods doesn’t make them an exception. However, knowing about these potential problems could help AI project teams to be proactive about avoiding them whenever possible.

Technically the underlying optimization development of online commerce forced problem is either NP or P-space businesses to question the week-plus supply-chain complete (depending on the details of the planning cycles that had been domain). Furthermore, the problem mixes the norm. Finally, the year 2000 (Y2K) a dozen or so classic optimization problems problem caused an across-the-board from AI and operations research (OR), replacement of enterprise software, allowing and much of the expected savings from many businesses to update their global supply-chain optimization are lost if approach to supply-chain planning. The end result of all of these factors was This article describes our experience a huge upswing in demand for supplychain from four years of solving supply-chain planning tools from i2 Technologies planning and optimization problems and other vendors. When I joined i2 in across industries, and some of the lessons 1996 as optimization architect, the company we learned.

These sorts of architectures were very much in vogue at the time, and the company wanted its own, proprietary technology. We started thinking about programming languages for the agents and the kinds of knowledge representation and reasoning that would be required. We spent a lot of time and money flying from London to the U.S. West Coast, talking to patent lawyers. It transpired that the architecture, its decision-making and action models, were completely inappropriate for the problem at hand. By the time we realized we should have been focusing on basic software engineering, quality assurance, and end-user requirements, the company had burned out much of the goodwill--and most of the funds--of our investors.

Extempo Systems, Inc. was founded in 1995 to commercialize intelligent characters. Our team built innovative software and novel applications for several markets. We had some early-adopting customers during the Internet boom, but the company was not quite able to survive the significant downturn in corporate IT spending when the bubble burst. In 2004, Extempo ceased operations and was formally liquidated. Although our commercial venture failed, we learned a lot, had fun, and are trying again with a new company. To others who aspire to commercialize their AI technology, I say: ";;Take a chance!";;

've spent over 20 years implementing algorithms for Boolean minimization, with particular application algorithms, dozens of conventional and exotic data structures, over two dozen programming languages and styles, and almost every level of implementation, from reconfigurable silicon through customized instruction sets, conventional programming languages, and very high-level languages such as Mathematica. The huge, randomly generated CNF formulae used to study SAT phase transition have attracted many creative approaches (such as variants of unit propagation, differential equations, probabilistic moments, component connectivity, cutting planes, and so on). However, I've learned one thing about the nature of Boolean minimization that seems obvious now. No matter how clever an algorithm is, no matter how extensively the structure of a problem is analyzed, no matter how much adaptive learning and lemma caching is used, the most successful approach to the general Boolean minimization problem is to use the simplest possible algorithm--brute force applied to lexicographically sorted formulae and implemented in reconfigurable silicon. For the general case, complexity cannot be finessed by any degree of cleverness.

In fact, DARPA, which funded the project, ways. Elves) (Scerri, Pynadath, and Tambe 2002; Finally, we will present some lessons Pynadath and Tambe 2003) and required learned and recent research that was motivated detailed information about the calendars by our experiences in deploying the of people using the system. Thus, we decided to deploy a new application of the Electric The Travel Elves introduced two major Elves, called the Travel Elves. This application advantages over traditional approaches to appeared to be ideal for wider deployment travel planning. First, the Travel Elves provided since it could be hosted entirely outside an interactive approach to making an organization and communication travel plans in which all of the data could be performed over wireless devices, required to make informed choices is such as cellular telephones. For example, when The mission of the Travel Elves (Ambite deciding whether to park at the airport or et al. 2002, Knoblock 2004) was to facilitate take a taxi, the system compares the cost planning a trip and to ensure that the of parking and the cost of a taxi given other resulting travel plan would execute selections, such as the airport, the specific smoothly. Initial deployment of the Travel parking lot, and the starting location Elves at DARPA went smoothly.