Scientists have revealed a way to make the brains of people more intelligent, and to study how we learn.
In a paper published in Nature, the team of scientists from MIT and Stanford describe a new system that they believe can “read” the mind of people, by using a “smart machine” to process and store “mental representations”.
The “smart machines” in question are “cognitive computers” that run in the cloud and have a set of neural nets.
When a person says a word, the computer uses its “learning engine” to learn the meaning of the word.
These computers have a “learning tree” that allows them to predict words that the person may have used in the past.
“If you say a word and you see it in your memory, that word was probably in your brain,” says Andrew Wiebe, a professor at MIT and one of the paper’s co-authors.
“We’re looking at how the brain processes these representations.”
The system is a “deep neural network” that learns to use “big data” to “learn” about the world, using neural nets and machine learning to do so.
This is where the brain becomes more complex than a single neuron.
“It’s kind of like the brain in a blender,” says Wieb.
“There are neurons in there, and you’re doing the same thing with this big data set.
It’s very, very, complicated.
So it’s the brain that we are trying to understand.”
Wiebe says the system’s capabilities will be very useful to medical researchers who are working with people with mental health conditions.
“We’re actually looking at a brain that is going to help us develop drugs and treatments for this,” he says.
“But we want to use this technology to study other brain systems in order to find out if we can treat this disorder.”
The team behind the new system, led by Andrew Wiese, PhD, and a team of researchers at Stanford, say their “smart computer” could be used to help doctors understand mental illnesses, or even to help “diagnose” people in the future.
Wieb says the idea is to create a new generation of machines that can “understand” how people work and what their needs are.
“The idea is we’re looking to go into the brain and say, what are the basic neural networks that are the basis of how you think and how you’re going to think about this?” he says, adding that the “big picture” of a person’s mental state is something that’s “really hard to explain.”
“If I say, I’m in this kind of a mental illness and I want to be better, I want you to see what that’s like.”
“What is it like to have a bad day, to have these negative thoughts and bad emotions?
What’s the cause?””
We need to understand how the system is developing these representations,” he explains.”
So if we have a machine that can understand that, it’s going to be able to help our doctors better understand what’s going on.””
The system itself is very complex.
And there are things that we don’t understand.
But we’re hoping that it can help us understand how to build these machines that understand how these representations are being built.”