Pieter Abbeel: Deep Reinforcement Learning | Lex Fridman Podcast #10

preview_player
Показать описание

Рекомендации по теме
Комментарии
Автор

After spending good 2 years on learning calculus, linear algebra, probability, python and finally deep learning (btw thanks to Andrew Ng) I'm starting to dive deep into DRL. Currently going through David Silver's lectures and RL bootcamp. Really like how Pieter explaines stuff there! Also, I was trying to start AI long time ago, but only Lex Fridman's DL intro lectures got my true attention. Simply and clearly explained and got me motivated to take this fun journey since then. It is so good to have you on this planet guys! :D

Trackman
Автор

This is an amazing interview: bunch of good questions and very sharp answers. Thanks to Pieter and Thank you Lex.

josedelgado
Автор

I think I am addicted to this series...
Thanks Lex!

mrghulamrasool
Автор

Another excellent conversation Lex! Love the deep question about happiness .Thanks very much!

supersnowva
Автор

"Love is the objective function..." Lex, you should copyright that. I want the T-shirt!

egparker
Автор

I learn so much from this channel that is entirely outside of my I'm a tattoo artist. This is brilliant. AI is very misunderstood, maybe even a misnomer, thanks to mainstream media and movies. Good questions and great answers. Still not married, btw...

voodooladyink
Автор

I like Peter and Levin's course. Thanks a lot to UCB and their devotion of open course. It helps get into robots and rl field.

perfumedsea
Автор

Awesome video! I am more amazed that this is Lex 10th video. It only has 62k views and 49 comments. As a software engineer, these are my favorite type of videos.

IvanGarcia-cxjm
Автор

Hey Lex, Thank you for creating and sharing these videos. 🙏🙏🙏🙏

taesiri
Автор

Thanks as always for uploading Lex. I'd never heard of Pieter Abbeel before but he seems pretty cool! Really interesting stuff.

lewisb
Автор

My first comment on youtube ever. Just wanted to thank you for all this great content. Please keep it up. :)

heinrichheidenreich
Автор

Lex ..I will keep this in my classic library

kparag
Автор

Great episode with great questions and answers.

tzjtjktzjtzjztjztj
Автор

Excellent video! I've been trying to build a couple RL robots and he addressed most of the problems I've run into. Now to read lots more papers.

praeclarum
Автор

Another great interview, thank you sir!

Lagruell
Автор

Great job Lex !!!!
I wonder,
Can we say Robots are conscious only after they truly experience the time ?

kirankolisetti
Автор

Why did I suspect if this was an AI and not really Lex!! (The intro)

VebitoTetseo_gPlus
Автор

Omg this is fucking dope. Good job Lex!
I still remember taking Berkeley's CS188 from Pieter, and learnt SO much about Artificial Intelligence.

ngocha
Автор

unsupervised learning is way to go for achieving AGI

bttmndl
Автор

Both Tesla and Google have a certain type of unit test. Google uses simulators a lot with all the sensor data they have from the real world. And as I understand it Tesla runs the new software in parallel on real cars so you can see what it would do.

autohmae