Supported by AITO. These pages are currently maintained by Alex Potanin |
|
Self-Enforcing Automatic Rational Contracts between Computers
University of Cambridge Computer Laboratory 15 JJ Thomson Avenue Cambridge CB3 0FD UK tel: fax: +44 (0)1223 334678 Brian.Shand@cl.cam.ac.uk Keywords:distribution frameworks language concepts meta architectures
Abstract:My area of interest is distributed systems --- my masters thesis developed a distributed object system for computer vision. I am now exploring applications of intelligent distributed agents, to both perform and manage distributed computerised tasks.Machine learning environments typically focus on concentrating as much intelligence as possible in a single agent. My thesis instead attempts to show how self-enforcing automatic rational contracts between computers can produce autonomous yet accountable computer systems, through the use of a realistic virtual economy. This thesis draws on techniques of distributed systems, machine intelligence and automatic programming, semantics and operating systems. The result would be a society of intelligent agents, solving problems quickly and efficiently in a competitive environment. I am also keen to explore tighter integration between language and middleware, to make this new environment directly accessible to programmers. On the one hand, this entails a natural object-oriented representation of concepts such as accounting policies, preferably in a constrained language accessible to the programmer, such as a Java or Python subset. On the other hand, new languages will be need to represent new types of behaviour, to make concepts such as composite events useful for programmers. Only by combining system frameworks, object-oriented middleware and languages can these new techniques be proven, used and shared in the computer science community.
HTML3 NO JAVA
Last modified on Mon Aug 15 14:59:24 2005 |