Catálogo de publicaciones - libros

Modal logics of strategic ability usually focus on capturing what it means for an agent to have a feasible strategy that brings about some property. While there is a general agreement on abilities in scenarios where agents have perfect information, the right semantics for ability under incomplete information is still debated upon. Epistemic Temporal Strategic Logic, an offspring of this debate, can be treated as a logic that captures properties of agents’ rational play.

In this paper, we provide a semantics of that is more compact and comprehensible than the one presented in the original paper by van Otterloo and Jonker. Second, we use to show that a rational player knows that he will succeed if, and only if, he knows how to play to succeed– while the same is not true for rational coalitions of players.

Many computer programs today show skills that appear to rival those of outstanding human consultants. However, while each such program does certain things well, it is helpless at doing anything else. Why do our present-day programs lack the versatility and resourcefulness that a typical person shows? Clearly, those programs are deficient in both commonsense knowledge and commonsense reasoning. I’ll argue that this has happened because the field of AI has evolved in a backwards direction, as compared with how a typical person develops-and that this is because our AI programmers have not appreciated the importance of making their system able to use more ’reflective’ ways to think.

I first heard about AI in 1969 from Max Clowes, then an AI vision researcher, when I was a philosophy lecturer at Sussex University (with a background in mathematics and physics). Gradually I came to realise that the best way to make progress in most areas of philosophy (e.g. philosophy of mind, epistemology, philosophy of language, philosophy of science, philosophy of mathematics, and probably even aesthetics) was to do AI.

The talk summarizes the beginnings of AI as a discipline in Germany until around 1982. This includes the formation of a community in 1975 by establishing a German AI conference series, a quarterly newsletter, and a representative body for AI within the GI, all of which is still in existence and thriving today. It also includes the embedding of the national activities within the international AI community by organizing the first "ECAI" conference as an AISB/GI conference in Hamburg, the foundation of ECCAI, the European AI umbrella society, and many other initiatives of German AI researchers.

This forthcoming book, jointly written by Corinna Elsenbroich (history and philosophy of science) and Jörg Siekmann (one of the founders of AI in Germany) explores the history of the establishment of AI in Germany with respect to two central themes:

(A) The pattern of development of a high-tech academic discipline (like AI) with substantial industrial and economic potential, follows an interesting, but different pattern from the well known and much older research subjects.

(B) The historical development of AI should not be viewed in isolation, but within the context of the so called ”Third Industrial Revolution”, i.e. the age of information processing and the computer, with all its societal consequences, such as economic globalization as well as the global information on the web.

Natural language understanding is one the most challenging goals of artificial intelligence. Since almost everyone speaks and understands a language, the development of natural language systems allows the average person to interact with computer systems anytime and anywhere without special skills or training, using common devices such as a cell phone. Full natural language understanding is AI-complete, in other words it requires solutions to all other core AI problems like knowledge representation, reasoning, vision, learning, and action planning. Nevertheless, after three decades of intensive and successful research, every day millions of users experience human language technology by calling directory assistance, getting train table or account information, dictating an SMS or a patient record, or telling a navigation system their destination. Human language technology has grown from an esoteric research area, 30 years ago, to a multi-billion euro market with a total revenue of more than two billion euro just for spoken dialog systems.

The “fifth decade of Artificial Intelligence”, the period from 1996 to 2006, has been a decade of success for AI. Computers beat the residing world chess champion. AI companies like Google and Yahoo! have been taking over the world. And AI research changed forever the nature of other scientific fields, like biology and cognitive science. So how could this happen?

In my humble opinion, the fifth decade of AI has been the decade of data and statistics. Data has been available for many years, but somehow the amount of available data exploded in the past decade. The advent of the World Wide Web made huge numbers of documents, images, and videos available online. Hundreds of AI researchers refocused their energy on the Web. AI systems were developed for learning people’s browsing patterns, for raking online music, for finding and parsing job advertisements, and for making search more effective. Out of this grew a billion-dollar industry, with Google being the most significant example.

The history of intelligent robotics started with Shakey developed at SRI in 1965. Shakey provided us several important research issues and we, robotics researchers, focused on the fundamental issues for making it more intelligent, such as Artificial Intelligence, Computer Vision, and Language Recognition. After Shakey, we have spent 40 years and developed humanoids as new Shakeys by using the developed technologies. The humanoids provide us new important research issues as Shakey did. The research issues lies on interdisciplinary areas among Robotics, Cognitive Science, Neuroscience, and Social Science.

During the last few days I learned a little about artificial intelligence and it was fascinating. I read an heard about things like machine learning and the idea of artificial emotions as the basis of artificial intelligence, which I had considered to be deepest science fiction some weeks before. In fact I just touched the surface, but the glimpse behind the curtain, that I got, left me stunned and made me think about the future. I thought how life would be in times, when you have to talk of intelligent artificial beings or even individuals instead of computer programs, because they seem to be living.