The neural network used by Uli Grasemann, a graduate student and his adviser, Professor Risto Miikkulainen at the University of Texas-Austin, is called DISCERN. Designed by Miikkulainen, DISCERN is able to learn natural language. The scientists taught it a series of simple stories, teaching it to store information as relationships between words and sentences – similar to the way a human being would learn a story.
“With neural networks, you basically train them by showing them examples, over and over and over again,” says Grasemann. “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.”
In order to model hyper learning, the scientists accelerated DISCERN’s rate of learning – so it was assimilating words at a faster rate, and it was not ignoring as much noise in the data. 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 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.
Telling the computer to “forget less” was akin to flooding the system with dopamine, confounding its ability to discern relationships between words, sentences and events.
“Information processing in neural networks tends to be like information processing in the human brain in many ways,” says Grasemann. “So the hope was that it would also break down in similar ways. And it did.”
The experiment doesn’t prove the hyper learning hypothesis, though it does strengthen their claims. It also shows that neural networks can be a useful analogue for the information-processing centers of the brain.
The researchers published their crazed computer findings in the journal Biological Psychiatry in April.