Sleep as a Specific Computational Process

We spend a third of our lives asleep, and science has long known sleep is essential for memory. But new theories go further: sleep doesn't just consolidate what you learned — it actively edits it, pruning away the noise and preserving what's most connected and meaningful.

We've built a formal model of this process, and it turns out the brain may be solving the same kind of optimisation problem that underlies compression in information systems.

What that means for how we think about learning, rest, and cognitive wellbeing is only beginning to come into focus.

Read the Paper

Look through the deck

Previous
Previous

We Know How to Make AI Safer. So Why Don't We?

Next
Next

AI Safety: LiberalArts Approach