As we keep edging towards Skynet self sufficient artificial intelligence, MIT have taken a considerable step closer with their take on a Neuronal chip, mimicking the function of a synapse in the brain.
We heard of DARPA and IBM creating a neurosynaptic chip a while back; but MIT are the first out the gates with the first physical evidence we've seen of the technology. In its capabilities of modelling specific communications among neurons, the chip could help further brain study and aid in the creation of neural prosthetic devices like artificial retinas.
“We now have a way to capture each and every ionic process that’s going on in a neuron,” said Chi-Sang Poon, a research scientist in the Harvard-MIT Division of Health Sciences and Technology and author of a paper describing the new chip.
The MIT Neuronal chip contains 400 transistors, all which act simultaneously to mimic the activity of a single brain synapse, which is the gap between which your neural and non-neural cells signal actions and thoughts, creating circuits between your mind and central nervous system. The chip works through recreating these actions through a tweak in chip transistors which allows MIT engineers to mimic a continuous flow, rather than the tradition on-off architecture common to other chips.
The brain contains around 100 billion neurons, connected in numerous ways through, performing many more hundreds of billions of transmissions per second. So for one synapse to be recreated is still a long ways away from a full AI recreation. But it's one step towards technology with self-awareness and capabilities beyond what anybody can probably comprehend thus far.