I have realized a long time ago that if we want to be able to produce systems that would have the same intellectual aptitudes compared to what we can find in the Nature, those systems would need to have similar characteristics, similar ways of working.

As of today, we do not know too much about Intelligence and Artificial Intelligence is mainly based on concepts that already exist in the Nature: Agent Based Models, Neuronal Networks, etc... I think we could say that what we try to do is understand Life and mimic it.

Based on this reflection, I identified the need for a software stack that would give us the ability to develop and reuse pieces of software simulating Life. This software stack would give us the ability to create autonomous entities, which would communicate with each other, and would also give us the ability to aggregate them forming other entities at a higher level of abstraction and probably run in parallel millions of them.

At that time and it is still probably true today, it was difficult, may be even impossible, to develop systems being able to run millions of agents on a single computer. Then naturally came to me the idea of running the agents on a distributed computer. Quickly enough, I realized that the effort to implement such a stack on a distributed computer is too important for a single person, mainly due the complexity of implementing all the layers needed to virtually assemble, manage and run a network of computers. I realized as well that the system would potentially face performance issues and the most important of all, the near impossibility to create a simple API.

If you observe what happened in the IT industry the last 20 years and if you apply the same evolution pattern to what could happen the next 20 years, it is not difficult to imagine that computers will be able to run simultaneously millions of threads and they would probably embed hundreds of cores, with probably some group of cores dedicated to specific functions and some others used for generic purposes.

Back early in 2005, I then started to implement a new software stack in Java called Artificial Life Framework, which executes all agents within the same Operating System process and which I hope could be used to simulate anything existing in real Life.

This year is an important year for me, I released the first version of what I think could help developers all around the world to develop systems that mimic Life and my hope is that in few years time, we will be able to leverage the work done to create intelligent system.