High Signal Podcasts Evidence ledger
Method
Browse
← Back to evidence

Evidence receipt / belief

Published · transcript-backed

Edward Gibson: belief

17 Apr 2024 Lex Fridman Podcast #426 – Edward Gibson: Human Language, Psycholinguistics, Syntax, Grammar & LLMs

“I agree with you that it would be nice to figure out… So, there’s some way to figure out the cost.”

— Edward Gibson

Source trail

Everything needed to verify it.

Speaker
Edward Gibson
Attribution
Verified speaker
Claim type
belief
Recorded
17 Apr 2024
Publisher
Lex Fridman Podcast

Transcript context

…So then you can somehow in different ways convert it to a number. I wonder if there’s a beautiful equation connecting cognitive costs and length of dependency. E equals MC squared kind of thing. It’s complicated, but probably it’s doable. I would guess it’s doable. I tried to do that a while ago, and I was reasonably successful, but for some reason I stopped working on that. I agree with you that it would be nice to figure out… So, there’s some way to figure out the cost. It’s complicated. Another issue you raised before was how do you measure distance? Is it words? It probably isn’t, is part of the problem, is that some words matter than more than others meaning nouns might matter. And then it maybe depends on which kind of noun. Is it a noun we’ve already introduced or a noun that’s already been mentioned? Is it a pronoun versus a name? All these things probably matter. So, probably the simplest thing to do is just like, oh, let’s forget about all that and just think about words or morphemes. For sure. But there might be some insight in the kind of function that fits the data. Meaning like quadratic… What-…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.

Search evidence