Scientists Induce Schizophrenia in a Computerized Brain


In an attempt to better understand the human brain, scientists at The University of Texas and Yale University used a virtual neural network to create schizophrenic thinking in a computer model. Their study showed that computer networks that "hyperlearn," or are unable to forget or ignore as much as normal, show symptoms of schizophrenia. This hyperlearning hypothesis is favored by some scientists as an explanation for schizophrenia, suggesting that schizophrenic brains lose the ability to identify what's meaningful because they are overwhelmed by too much information.

Scientists taught the neural network, called DISCERN, a series of simple stories that DISCERN incorporated into its memory in a way similar to how the human brain stores information. They then modeled hyperlearning by increasing the network's learning rate, which simulated an excessive dopamine release in the brain and kept DISCERN from forgetting as much as it normally would.

When DISCERN was taught using hyperlearning, it began showing abnormalities in language that resembled schizophrenia. For example, according to the University of Texas new release, "DISCERN began putting itself at the center of fantastical, delusional stories that incorporated elements from other stories it had been told to recall. In one answer, for instance, DISCERN claimed responsibility for a terrorist bombing.

In another instance, DISCERN began showing evidence of 'derailment' — replying to requests for a specific memory with a jumble of dissociated sentences, abrupt digressions and constant leaps from the first- to the third-person and back again."

While it doesn't prove that the hyperlearning hypothesis is right, this new study does offer support for the hypothesis by showing that hyperlearning can lead to schizophrenic thinking (at least in computers). The successful use of a computer model of the human brain is also an important outcome of the study, and is a tool that could prove useful for future research.

Read more: