YouSaid · the spoken record

Balaji Srinivasan

lines on the record: 605
first: 2022-10-20
most recent: 2022-10-20
sittings or episodes: 1
sources: podcast

Every line below is reproduced as it was said and linked to the record it came from. Nothing here is summarised or generated. Directory · Search · Corrections

“Yeah, people will typically quote Cordell's incompleteness for such a question. And yeah, there are things that are provably unknowable or provably unprovable. But I think you can often get an approximate solution. The Hilbert's problems like we will know, we must know. At least we should know that we can know, push to get at least an approximate solution, push to know that we can't know. At least we push back that darkness enough so that we have lit up that corner of the intellectual universe.”
2022-10-20 · Lex Fridman Podcast · #331 – Balaji Srinivasan: How to Fix Government, Twitter, Science, and the FDA · IDENTIFIED FROM THE TRANSCRIPT · source
“Than vice versa to some extent that's happening in neural networks now, as you're aware, right? And I think that's, so just something that's true and that works, you know, if we don't know yet, that's amazing and that pulls us forward. So I do think that the limits are more of a bug than a feature.”
2022-10-20 · Lex Fridman Podcast · #331 – Balaji Srinivasan: How to Fix Government, Twitter, Science, and the FDA · IDENTIFIED FROM THE TRANSCRIPT · source
“Think mostly a bug. I admire Romanogen. I admire Feynman. I admire these great mathematicians and physicists who were just able to see things that others couldn't. And just by writing it down, that's a leap forward. People talk about it's not the idea, it's execution, but that's for trivial ideas, for great ideas for Maxwell's equations or Newton's laws or quantum electrodynamics or some of Romanian's identities. That really does bring us forward, especially when you can check them. You don't know how they work, right? You have the phenomenological, but you don't have the theory underneath it. And then that stimulates the advancement of theory to figure out why is this thing actually working? And that's actually, you know, Statmac arose in part from the kind of phenomenological studies that were basically being done where people are just getting steam engines and so on to work. And then they kind of abstracted out thermodynamics and so on from that, right? So the practice led the theory, right?”
2022-10-20 · Lex Fridman Podcast · #331 – Balaji Srinivasan: How to Fix Government, Twitter, Science, and the FDA · IDENTIFIED FROM THE TRANSCRIPT · source
“Yet, the rule is actually very simple. And so the thing I think about a lot is just how many patterns in life are we just like these rats and we're trapped in a prime number maze. And if we had just a little bit more cogitation, if we had a little bit more cognitive ability, a little bit more whether it's brain machine interface or just better physics, we could just figure out the next step in that primary mason. We could just see it. We could just see the grid, right? And that's what I think about like that's a big thing that drives me. Figuring out how do we actually conceive, understand that prime number maze that we're living in.”
2022-10-20 · Lex Fridman Podcast · #331 – Balaji Srinivasan: How to Fix Government, Twitter, Science, and the FDA · IDENTIFIED FROM THE TRANSCRIPT · source
“All right, great. A rat can be trained to turn at every even number or every third number in a maze to get some cheese. But evidently it can't be trained to turn it prime numbers. Two, three, five, seven, and then eleven and so on and so forth. That's just too abstract. And frankly, if most humans were dropped into a prime number maze, they probably wouldn't be able to figure it out either. They'd have to start counting and so it actually be pretty difficult to figure out what.”
2022-10-20 · Lex Fridman Podcast · #331 – Balaji Srinivasan: How to Fix Government, Twitter, Science, and the FDA · IDENTIFIED FROM THE TRANSCRIPT · source