Computational wizards have been trying to copy the behavior and structure of the human brain in all sorts of ways. In one recent stab, a group created a digital model that not only reproduced aspects of complex brain behavior, it even made similar mistakes.
This virtual model is called “Spaun”–for Semantic Pointer Architecture Unified Framework. The digital “brain” can receive visual cues and sketch responses to them with a mechanical arm. It can do basic tasks like complete lists of numbers or solve simple arithmetic problems—tasks regular people encounter in IQ tests. Surprisingly, the model even picked up on bizarre brain behavior, like remembering the first and last numbers of a list better than other members.
The Spaun brain simulation involves 2.5 million virtual neurons. That’s a mere handful compared to the human brain’s 86 billion neurons, but that’s part of the point. The goal of the Spaun team is not to replicate physiology neuron-for-neuron, but rather to reproduce complex behavior. In contrast, other big brain modeling groups like the Blue Brain Project, seek to achieve a high level of biological accuracy with as many neurons as possible, with the hope that complex behavior will eventually follow.
The Spaun team explains in their paper published last week in Science:
…simulating a complex brain alone does not address one of the central challenges for neuroscience: explaining how complex brain activity generates complex behavior. In contrast, we present here a spiking neuron model of 2.5 million neurons that is centrally directed to bridging the brain-behavior gap.
Nature News has a great explanation of how Spaun makes connections, in some ways mirroring the working of the brain itself:
The computing cells are divided into groups, corresponding to specific parts of the brain that process images, control movements and store short-term memories. These regions are wired together in a realistic way, and even respond to inputs that mimic the action of neurotransmitters.
For all its cleverness, Spaun does come with several shortcomings, not surprising given its scale. But it does have its place among models that seek to understanding the brain–after all, given the complexity of that project, it’s all hands on deck.