The monkey and the machine: a dual process theory – The sideways view

I think that it’s worth having a simple picture of my own brain; this post presents my working model, which I’ve found extremely useful.

I think that it’s worth having a simple picture of my own brain; this post presents my working model, which I’ve found extremely useful. I. The model The monkey is a fast subconscious process, something like an RNN running at 100 Hz with a billion-dimensional hidden state. From a computational perspective, the monkey is where the action is. The monkey doesn’t think in words. It gets an input, and reflexively produces an output. We have no direct introspective access to anything happening inside the monkey. The monkey is trained by something like reinforcement learning, with a

saved by

• Uzay Girit
• Vincent Cheng

related reading

• A Toy Environment For Exploring Reasoning About Reward — LessWronglesswrong.com
• Deep Reinforcement Learning: Pong from Pixelskarpathy.github.io
• A Crash Course in the Neuroscience of Human Motivation — LessWronglesswrong.com
• pdfopenreview.net
• God Help Us, Let's Try To Understand The Paper On AI Monosemanticityastralcodexten.com
• So how does learning work?cs.mcgill.ca
• Embedded Agency (full-text version) — LessWronglesswrong.com
• Conflicts between emotional schemas often involve internal coercion — LessWronglesswrong.com
• Models Don't "Get Reward" — LessWronglesswrong.com
• Deep Reinforcement Learning: Pong from Pixelskarpathy.github.io
• How training-gamers might function (and win)blog.redwoodresearch.org
• Reward Is Not Enough — LessWronglesswrong.com