Dr. Michael Kilgard

That all came on with the Industrial Revolution.

A thunderclap was like a loud sound.

Nowadays, you can go to a concert and be at 120 decibels, 130 decibels for hours at a time.

Most people, interestingly, don't get tinnitus.

We see the same thing in the rats.

When we produce noise trauma, I can make everybody deaf.

That's easy.

But only half of them show signs of tinnitus.

So the way the brain rewires itself depends upon the specifics of creating that positive feedback loop inside the brain.

Similar things happen.

Charlie Gilbert did these beautiful experiments with the retina.

We make dual lesions in the retina, and all the neurons will go jump over to some crazy place.

We now have computational models where we understand the neurons have no inputs.

They will tend to do dumb things.

They don't tend to do smart things because you didn't tend to have simultaneously paired retinal lesions.

So trusting the network to do the right thing is a good idea.

Brains are smart.

But in the case of pathology, by definition, it's not doing the right thing or it would have fixed it.

So the same thing is happening with cancer.

We get cancers all the time, and we fix them.