The Blue Brain project and the technological singularity

The technological singularity it can be reached with the artificial intelligence (AI), the biology, the nanotechnology, the biotechnology or the bioinformatics. Here I'm going to talk about AI.

Some important scientist at the field of AI have made predictions about the evolution of the computational brains leaning in the Moore’s law. Next you can see a pair of graphs of these predictions, the first of Hans Moravec and the second of Raymond Kurzweil.

According to this the future will can be how the proposed in Matrix or Terminator. Instead, if the machines will not rebel to us, we could arrive to be immortals and all the problems of the humanity would be solved. But are being performed these predictions?

The fact is that we have not a complete model of the behaviour of the neurons, nor of the cognitive process of high level that emerge from that, nor we have got a clear idea of the processing capacity of the brain. For these reasons is quite questionable that this predictions would be achieved in a short-term.

At present there are infinity of techniques at artificial intelligence to get intelligent behaviours (neural nets, fuzzy logic, evolutionary computation, etc.) These techniques are good resolving determinate problems, but none can be applied with success to any situation. Although maybe would can made an architecture joining these techniques to achieve a behaviour seemed to an intelligent entity.

To create intelligences that serve for any problem, the most practical would be find a unified theory of the artificial intelligence, an algorithm that allows simulate any intelligent behaviour of any living form. Probably it would by an algorithm that made elementary functions and joining instances of itself, an intelligence behaviour will emerge.

To follow this second approximation we have to do a process of reverse engineering of the brain, that is to say, know how work and emulate it in a computer. With this idea arises the project Blue Brain in 2005. In 2006 they finished a model of a cortical column of a rat, where, supposedly, would be the conscience and other superior functions of the brain. To execute only one cortical column (a human have a million of them) they need make use of all resources of the supercomputer IBM Blue Gene.

According to this we could think that we are quite far of simulate a complete brain of a rat, point where we should have arrived according to the graph displayed previously, but we have to consider that this model of the cortical column doesn’t simulate only the behaviour of the neurons, also how works showing it in 3D images, something that wouldn’t be necessary make in the practice in a IA. Furthermore the model represents every single detail, where can exists parts that wouldn’t be necessary in practice.

The 3D images of the Blue Brain Project, allow observe high level patterns that arises of the spontaneous organization of the neurons and its study could allow find the theory of everything in AI and achieve the technological singularity.

The researches of the project, in their predictions, match with the graphs when say that about the year 2020 (with an error of 5 years more or less), they will have a human artificial brain. In some years we will see if they are wrong or not.

Finally to make you think about this, here I leave a video of a robot that works with rat neurons.

See at YouTube