Is Consciousness a Theory of Mind?

4min Snip

00:00

Play full episode

Summary

Transcript

Episode notes

I would say I'm like agnostic about the notion of consciousness rather than convinced consciousness doesn't exist but it annoys me that there's this sort of presumption that humans have access to some higher, some metaphysically given notion of consciousness. Like again behaviorism is a theory of my as a philosophy of mine is you know full of weakness and counter arguments from at least 30 years of philosophy of mine literature. It just seems like we're throwing away too readily in the argument being made by the stochastic parent crowd the possibility that this is really all that's happening with humans right? That kind of anthropocentricism can't go unchallenged so I sometimes like to stir trouble

In episode 62 of The Gradient Podcast, Daniel Bashir speaks to Ed Grefenstette.

Ed is Head of Machine Learning at Cohere and an Honorary Professor at University College London. He previously held research scientist positions at Facebook AI Research and DeepMind, following a stint as co-founder and CTO of Dark Blue Labs. Before his time in industry, Ed worked at Oxford’s Department of Computer Science as a lecturer and Fulford Junior Research Fellow at Somerville College. Ed also received his MSc and DPhil from Oxford’s Computer Science Department.

Have suggestions for future podcast guests (or other feedback)? Let us know here!

Subscribe to The Gradient Podcast: Apple Podcasts | Spotify | Pocket Casts | RSSFollow The Gradient on Twitter

Outline:

* (00:00) Intro

* (02:18) The Ed Grefenstette Origin Story

* (08:15) Distributional semantics and Ed’s PhD research

* (14:30) Extending the distributional hypothesis, later Wittgenstein

* (18:00) Recovering parse trees in LMs, can LLMs understand communication and not just bare language?

* (23:15) LMs capture something about pragmatics, proxies for grounding and pragmatics

* (25:00) Human-in-the-loop training and RLHF—what is the essential differentiator?

* (28:15) A convolutional neural network for modeling sentences, relationship to attention

* (34:20) Difficulty of constructing supervised learning datasets, benchmark-driven development