Physicists at Emory University have published research that shows how they turned a mouse’s brain activity into a predictive model. This could be a huge discovery for artificial neural networks. The team of researchers used machine learning to grow a theoretical mouse brain.

Emory University press release reads:

The dynamics of the neural activity of a mouse brain behave in a peculiar, unexpected way that can be theoretically modeled without any fine tuning.

It means that they can observe a mouse’s brain activity in real-time; however, the number of neuronal interactions is too high for them to quantify each one of them – even with the help of an AI. So the team is using something similar to math to measure these interactions.

The research is based on the theory that neurons in the brain exist in equilibrium. And, not all neurons do the same thing, but also don’t move around randomly. The researchers ran a bunch of tests on mice to establish brain data.

Then the team started developing a simplified model to predict neuron interactions using experimental data as a target. Since the research is in its early phase, scientists have used mice brains because they are not very different from humans. If a working Ai model can reduce what’s happening in a mouse’s brain, it’s possible to be used on a human brain in the future.

The new AI model looks promising, but its actual usefulness is yet to be determined. The tech and AI industries are also at an unpredictable point where hardware solutions and sophisticated software are growing away from one another. 

Still, if we stay optimistic, this could also be the beginning of artificial general intelligence (AGI) – machines that have a brain of their own. But to get there, we will have to start with models capable of imitating organic neural nets.