Yang-Hui He

speaker
1,500 appearances 1 recordings 1 series first heard Jan 2025 last heard Jan 2025

Yang-Hui He’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months
No recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.

Appearances

newest first · ▶ plays the moment
And you know, Hardy
in in this in this beautiful in his beautiful apology says, you know, what mathematicians do is essentially we are um pattern recognizers.
That's that's probably the best definition of what mathematics is, or is is that is a study of patterns.
finding regularity in patterns.
And in fact, you know, if there's one thing that AI can do better than us.
is pattern detection.
Because, you know, we we evolved in being able to detect patterns in three dimensions and no more.
So in this sense, if you have the right representation of data, um, you're sure that AI can do better than that.
I mean, of you know, generate a lot of stuff, but filtering out what is better is is a very interesting problem in and of itself.
So let's try to do one.
I mean there are various ways to do this representation.
One one way you can do it
is to do what a problem which is m
maybe best fit for for an AI system, which is binary classification of binary vectors.
So what you do is what you know, sequence prediction is kind of difficult.
So what one thing you can do is just take this infinite sequence and just take, say, a window of a hundred, a thousand words, fixed window size, and then label it with a the one immediately outside the window and then shift, label, shift, label.
So and then
Then you can generate a lot of training training data this way.
Uh-huh.
So for this sequence, you've I think I've just taken here, you know, whatever the sequence is and and just with a fixed window size and with this label.
Showing 681–700 of 1,500 · page 35 of 75 ← Previous Next →