Scientists Attempt to Induce Schizophrenia on a Computer
A recent study by researchers at the University of Texas in Austin and Yale University sought to create the thinking of a schizophrenic mind on a computer using a virtual neural network. Their work is based on the so-called hyperlearning theory of schizophrenia, which holds that the disease springs from an inability to forget or ignore non-essential information.
In their work, the research team taught a series of stories to a computer model they call DISCERN. Using natural language processing, the computer maps out the stories in a manner similar to the human brain. In their model, a simulated dopamine release is used to mark significant information as DISCERN learns the stories. To model hyperlearning, the team increased the dopamine releases. This meant DISCERN “forgot” less, and perceived more information as “important.”
When asked to recount the stories, the hyperlearning DISCERN produced bizarre and delusional narratives from the input information. From Science Blog:
After being re-trained with the elevated learning rate, 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 intriguing, and eerily similar to actual schizophrenia, the DISCERN study is not hard proof of the hyperlearning hypothesis. It is, after all, a simulation and the relevance of its output is interpreted by humans. But the innovative approach to the study, modeling a cause of a brain disorder and comparing the results to actual cases, is remarkable in its own right and could prove a powerful new tool for doctors.
(Science Blog via Slashdot, image via Akelei)
Have a tip we should know? tips@themarysue.com
