Paranoid Android: Simulating Schizophrenia on a Computer
"A mental breakdown in the relation between thought, emotion and behavior, leading to faulty perception, inappropriate actions and feelings, withdrawal from reality and personal relationships into fantasy and delusion, and a sense of mental fragmentation."
Sound like what your brain feels like after an Angry Birds marathon or an 8-hour binge of searching for the end of the Internet? Maybe. But this is actually the definition of schizophrenia.
Ubiquitous computing might send schizophrenic ripples through your frontal lobes whenever you log on, but researchers at the University of Texas at Austin are taking that a step further by actually using computers to simulate the mental disorder.
One theory about the cause of schizophrenia suggests that an afflicted brain remembers too many irrelevant things, due to an excessive release of dopamine. Overwhelmed by vast caches of facts, thoughts and memories all junk-piled in their heads, schizophrenics start processing them into delusional conclusions not based in reality.
Uli Grasemann, a graduate student in the University of Texas' Department of Computer Science, used a synthetic neural network to simulate these delusions, often called hyperlearning hypothesis.
Designed by his adviser, professor Risto Miikkulainen, Graseman used the network, dubbed DISCERN,to mimic the effects that different neurological dysfunction have on human language function.
Graseman and Miikkulainen began by programming DISCERN to remember a series of stories using a process similar to human memory.
"With neural networks, you basically train them by showing them examples, over and over and over again," Grasemann said in a university press release. "Every time you show it an example, you say, if this is the input, then this should be your output, and if this is the input, then that should be your output. You do it again and again thousands of times, and every time it adjusts a little bit more towards doing what you want. In the end, if you do it enough, the network has learned."
Next, they modified DISCERN to stop filtering out extraneous and irrelevant information. In other words, they programmed it to forget less, resulting in "fantastical, delusional stories" that spliced together unrelated stories. In one instance, DISCERN confessed to being the culprit of terrorist bombing.
Ralph Hoffman, professor of psychiatry at the Yale School of Medicine analyzed the behavior of the network and found it similar to schizophrenic behavior in humans.
Grasemann thinks the experiment shows great potential for using computers in neurological research.
"Information processing in neural networks tends to be like information processing in the human brain in many ways," he said. "So the hope was that it would also break down in similar ways. And it did."
Photo: Lisa Valder
Comments