Who is LEIA?

At present, LEIA is a one man project. I, Jeroen van Maanen, started it in 2003 as the next step in realizing an idea that occurred to me in 1995. That idea was:

Learning can be a goal in itself that surpasses and contains all subgoals that a system can have with respect to its environment

I have been fascinated with Artificial Intelligence (AI) since I read Gödel, Escher, Bach in 1986 (more than six year after its publication). However, I felt that AI tried to jump to the top of the stairs. Gradually I began to see that the mystery of intelligence is not as much the ability to solve problems, but the ability act with an inner sense of purpose.

During my time as a graduate student in the field of Computational Learning Theory at CWI I got the idea that is central to LEIA, but it was too early for me to compose a setting that could lead to a first step in realizing this idea.

Slowly over what turns out to be decades, I am working on a software library that implements the principle that I stumbled upon in my CWI years. Hopefully it will evolve to a level where it transcends the purely academic.

Some ideas:

• I would be happy to help write (even co-author) papers on the theoretical aspects of the Autonomous Learning principle
• Use the LEIA software to explore complex environments where no well-defined utility functions exist (yet), for example,
  • Bio-informatics
  • Space exploration
  • Industrial control where control strategies influence the nature of the utility function
  • Theory of games where the sum of the gains of the players need not be zero
• Use the LEIA software in simulation software to model autonomous agents
• Use the LEIA software in software games to model computer controlled players or 'bots.
• Extend the LEIA software to create interactive control systems for artificial limbs, integrating with nerve tissue of the wearer
  • (long term) These interactive systems that replace nerve tissue could replace damaged brain tissue
    • (really really long term) An external artificial learning system based on LEIA could be intergrated so closely with a human brain that over time it would become one big system of interdependent components where the artificial part would continue to be the same person when the biological part deteriorates: a vehicle for the mind

Interested? Contact me using the contact form.

Visit the LEIA project page on GitHub for project details.