The Society for the Study of Artificial Intelligence and the Simulation of Behaviour is proud to sponsor a discussion event hosted by the Royal Society of Edinburgh. The event is being co-organised by the SSAISB and the British Computer Society Specialist Group on AI (SGAI). For more events at the Royal Society of Edinburgh, please see: HERE. A poster for the event is HERE.

Title: Artificial Intelligence: In your Life Today

Date: Friday 5 August 2005 at 6pm

Chair:

Professor Alan Bundy, Professor of Artificial Intelligence, The University of Edinburgh, UK. Alan's homepage is HERE.

Speakers:

Professor Aaron Sloman, Professor of Artificial Intelligence and Cognitive Science, The University of Birmingham, UK. Aaron's homepage is HERE.

Professor Wolfgang Wahlster, Director and CEO of the German Research Centre for Artificial Intelligence DFKI, Saarbruecken, Germany and Professor of Computer Science at Saarland University, Germany. Wolfgang's homepage is HERE.

Synopsis:

Science fiction books and recent movies focus on intelligent robots, but Artificial Intelligence covers a much broader scope: we study human and artificial minds, aiming to produce intelligent computing systems which have a direct, beneficial impact on the lives of countless people. How do we build systems which can undertake intelligent activities? Where are these systems, and how can they change lives for the better?

Abstract for Prof. Sloman's talk:

Despite many exaggerated claims, AI is still decades away (some think it may be centuries away) from explaining or replicating, the competence of a nest building or hook-making bird, or the learning capabilities of a human toddler. There are several reasons for this. The main one is immense complexity of bodies, brains and minds produced by several billion years of biological evolution -- followed by extraordinary growth of competence produced by cultural evolution. Other reasons why progress has been slow include our inadequate understanding of what animals and children can do, and our inability to account for the information-processing involved in biological organisms. Moreover we are still far from being able to make bodies and body parts with the mechanical capabilities of animal bodies, including their manoeuvreability, their controllability and their power/weight ratios.

Certainly, over-optimistic predictions about how fast progress in AI would be made turned out to be completely false. But that does not show that AI has failed, as some claim. It merely shows that the people who made the predictions did not understand the problems. And it ignores the many successes in AI, both as science and as engineering. Despite those successes, AI is still a science in its infancy and if we analyse the problems with great care, and try to get people from different disciplines and different subfields working together instead of trying to solve all their problems in isolation, there is hope for sustained continuing progress on one of the greatest challenges of all time, understanding the variety of types of mind produced by evolution.

We can use that increased understanding to produce a succession of robots and other machines that approach human capabilities (although, as with all new technologies we shall have to take care that they are put to good, not bad, uses). We shall also acquire a deeper understanding of what we are and how we work than we have ever had before, and this will have profound effects on our approaches to education, counselling, and therapy.

Making progress requires some of the brightest young minds to choose to study the disciplines that need to be combined in this endeavour, including not only AI, but also philosophy, psychology, linguistics, neuroscience, biology, anthropology, ethology, engineering and, of course, mathematics. But students of these disciplines must not allow administrative and funding constraints, or narrow-minded teachers, to keep their work focused within the borders of specific disciplines or research groups. Progress requires our vision to be both broad and deep. It's very difficult. And very exciting.

Abstract for Prof. Wahlster's talk:

AI Entering Everyday Life: The Smart Car Example

Artificial Intelligence (AI) to the Hollywood generation is mostly about androids, humanoids and robots. It is about machines going out of control, replacing humanity and world domination. However, in the reality of the 21st century AI technology is not threatening, but becomes smoothly integrated into the fabric of everyday life. It adds knowledge and reasoning to existing applications, to make them more flexible, smarter, easier to use, and more sensitive to user behavior and changes in their environments. Maturing AI technologies have left the research labs and become ubiquitous to the point where they are almost invisible.

This brief talk will use the example of the smart car to illustrate this point. Speech recognition, speech synthesis and dialog understanding are used to enable the driver to control the air conditioning, to select music or to program the navigation system. The age and gender of the driver is detected by biometric voice analysis and the cockpit displays are automatically adapted to the selected user model. Semantic web services can be activated that find the lowest price gas station close to the current position of the car. Up-to-date information about historical landmarks and points of interest on the car's route is retrieved in a location-sensitive way from the Internet. Automated deduction is the basis for correctness proofs of safety-critical car components like ABS both on the hardware and software level. Pattern recognition is used to classify sensor signals and warn the driver e.g. about the risk of aquaplaning. Messages are sent via wireless ad-hoc networks to succeeding vehicles to warn their drivers of potential road hazards. The analysis of biosensors and the driving behavior is used to infer the cognitive and emotional state of the driver, so that the smart car can assist the driver in a personalized and situation-adaptive way. Various video clips will illustrate our joint work with DaimlerChrysler, BMW, Porsche, Audi, and Volkswagen on smart cars.

The industrial success gives lie to the provocative definition that AI stands for "Almost Implemented". In fact, AI is today routinely employed in so many areas of informatics, that it is fair to say that AI stands for Avant-garde Informatics, since it is always pushing informatics to its limits. Bold innovations always require equally bold imagination.

This talk is dedicated to the memory of the late Rob Milne, whose passion was transforming AI research results into applied technologies with a real effect in the real world.

Organisers

The late Dr. Rob Milne was a driving force behind this event. For a memorial web page, please see: HERE.

This event is being organised by Dr. Simon Colton from the AISB society. His web page is HERE.