Bart Kosko on noise and optimisation

Neural network and ‘fuzzy thinking’ researcher Bart Kosko is briefly interviewed in this month’s Wired where he argues that adding noise to a system, including the human one, may improve performance.

It reminded me of part of a colourful interview he did for the 1998 book Talking Nets: An Oral History of Neural Networks – a wonderful collection of personal memories from key scientists in artificial intelligence.

I like to ask researchers where they get their ideas. The only answer I’ve heard that makes sense is, “You vary your input if you want to vary your output.” Do lots of things. If you’ve gotta take drugs, take drugs. Take long walks, meditate, watch a lot of movies, learn a new language, read different books, argue the other side of the debate – anything you can to vary your stimuli.

And then you have to, as they say, “keep the ass in the seat.” You actually have to sit down and write. Do it in a disciplined way. I think if people have a certain minimal training in mathematics, the problem will take care of itself because neural networks are inherently interesting, and I believe they will stay interesting well into the next century.

The rest of Kosko’s Talking Nets interview covers topics as diverse as libertarian politics, cognitive maps, God, the mathematics of fuzzy systems, the economics of marijuana, organising neural network conferences and cryogenic nanobots.

Link to brief Kosko interview in Wired.
Link to Talking Nets book details.

UPDATE: Thanks to Daniel for finding the full Talking Nets interview on Google Books. You can read it here.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s