episode
stringlengths
45
100
text
stringlengths
1
528
timestamp_link
stringlengths
56
56
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
mind or the artificial mind?
https://karpathy.ai/lexicap/0004-large.html#00:41:28.880
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
You know, when I was an adolescent, I was reading a lot. And then I started reading
https://karpathy.ai/lexicap/0004-large.html#00:41:31.360
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
science fiction.
https://karpathy.ai/lexicap/0004-large.html#00:41:35.520
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
There you go.
https://karpathy.ai/lexicap/0004-large.html#00:41:36.520
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
That's it. That's where I got hooked. And then, you know, I had one of the first personal
https://karpathy.ai/lexicap/0004-large.html#00:41:37.520
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
computers and I got hooked in programming. And so it just, you know,
https://karpathy.ai/lexicap/0004-large.html#00:41:46.520
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
Start with fiction and then make it a reality.
https://karpathy.ai/lexicap/0004-large.html#00:41:52.680
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
That's right.
https://karpathy.ai/lexicap/0004-large.html#00:41:54.800
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
Yoshua, thank you so much for talking to me.
https://karpathy.ai/lexicap/0004-large.html#00:41:55.800
Yoshua Bengio: Deep Learning | Lex Fridman Podcast #4
My pleasure.
https://karpathy.ai/lexicap/0004-large.html#00:41:57.560
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
The following is a conversation with Vladimir Vapnik.
https://karpathy.ai/lexicap/0005-large.html#00:00:00.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
He's the co inventor of support vector machines,
https://karpathy.ai/lexicap/0005-large.html#00:00:03.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
support vector clustering, VC theory,
https://karpathy.ai/lexicap/0005-large.html#00:00:05.200
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and many foundational ideas in statistical learning.
https://karpathy.ai/lexicap/0005-large.html#00:00:07.840
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
He was born in the Soviet Union
https://karpathy.ai/lexicap/0005-large.html#00:00:11.120
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and worked at the Institute of Control Sciences in Moscow.
https://karpathy.ai/lexicap/0005-large.html#00:00:13.080
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Then in the United States, he worked at AT&T,
https://karpathy.ai/lexicap/0005-large.html#00:00:16.240
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
NEC Labs, Facebook Research,
https://karpathy.ai/lexicap/0005-large.html#00:00:19.280
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and now is a professor at Columbia University.
https://karpathy.ai/lexicap/0005-large.html#00:00:22.200
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
His work has been cited over 170,000 times.
https://karpathy.ai/lexicap/0005-large.html#00:00:25.880
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
He has some very interesting ideas
https://karpathy.ai/lexicap/0005-large.html#00:00:30.120
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
about artificial intelligence and the nature of learning,
https://karpathy.ai/lexicap/0005-large.html#00:00:31.800
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
especially on the limits of our current approaches
https://karpathy.ai/lexicap/0005-large.html#00:00:34.760
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and the open problems in the field.
https://karpathy.ai/lexicap/0005-large.html#00:00:37.560
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
This conversation is part of MIT course
https://karpathy.ai/lexicap/0005-large.html#00:00:40.360
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
on artificial general intelligence
https://karpathy.ai/lexicap/0005-large.html#00:00:42.440
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and the artificial intelligence podcast.
https://karpathy.ai/lexicap/0005-large.html#00:00:44.360
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
If you enjoy it, please subscribe on YouTube
https://karpathy.ai/lexicap/0005-large.html#00:00:46.800
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
or rate it on iTunes or your podcast provider of choice,
https://karpathy.ai/lexicap/0005-large.html#00:00:49.520
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
or simply connect with me on Twitter
https://karpathy.ai/lexicap/0005-large.html#00:00:52.960
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
or other social networks at Lex Friedman spelled F R I D.
https://karpathy.ai/lexicap/0005-large.html#00:00:55.240
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
And now here's my conversation with Vladimir Vapnik.
https://karpathy.ai/lexicap/0005-large.html#00:01:00.120
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Einstein famously said that God doesn't play dice.
https://karpathy.ai/lexicap/0005-large.html#00:01:04.760
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Yeah.
https://karpathy.ai/lexicap/0005-large.html#00:01:08.800
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
You have studied the world through the eyes of statistics.
https://karpathy.ai/lexicap/0005-large.html#00:01:09.920
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
So let me ask you in terms of the nature of reality,
https://karpathy.ai/lexicap/0005-large.html#00:01:12.800
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
fundamental nature of reality, does God play dice?
https://karpathy.ai/lexicap/0005-large.html#00:01:17.280
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
We don't know some factors.
https://karpathy.ai/lexicap/0005-large.html#00:01:21.320
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
And because we don't know some factors,
https://karpathy.ai/lexicap/0005-large.html#00:01:25.400
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
which could be important,
https://karpathy.ai/lexicap/0005-large.html#00:01:28.160
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
it looks like God plays dice.
https://karpathy.ai/lexicap/0005-large.html#00:01:30.520
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
But we should describe it.
https://karpathy.ai/lexicap/0005-large.html#00:01:35.040
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
In philosophy, they distinguish between two positions,
https://karpathy.ai/lexicap/0005-large.html#00:01:38.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
positions of instrumentalism,
https://karpathy.ai/lexicap/0005-large.html#00:01:42.080
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
where you're creating theory for prediction
https://karpathy.ai/lexicap/0005-large.html#00:01:44.920
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and position of realism,
https://karpathy.ai/lexicap/0005-large.html#00:01:48.720
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
where you're trying to understand what God did.
https://karpathy.ai/lexicap/0005-large.html#00:01:50.960
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Can you describe instrumentalism and realism a little bit?
https://karpathy.ai/lexicap/0005-large.html#00:01:54.640
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
For example, if you have some mechanical laws,
https://karpathy.ai/lexicap/0005-large.html#00:01:58.400
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
what is that?
https://karpathy.ai/lexicap/0005-large.html#00:02:04.200
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Is it law which is true always and everywhere?
https://karpathy.ai/lexicap/0005-large.html#00:02:06.280
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Or it is law which allow you to predict
https://karpathy.ai/lexicap/0005-large.html#00:02:11.480
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
position of moving element?
https://karpathy.ai/lexicap/0005-large.html#00:02:14.880
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
What you believe.
https://karpathy.ai/lexicap/0005-large.html#00:02:19.880
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
You believe that it is God's law,
https://karpathy.ai/lexicap/0005-large.html#00:02:23.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
that God created the world,
https://karpathy.ai/lexicap/0005-large.html#00:02:25.520
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
which obey to this physical law.
https://karpathy.ai/lexicap/0005-large.html#00:02:28.520
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Or it is just law for predictions.
https://karpathy.ai/lexicap/0005-large.html#00:02:33.200
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
And which one is instrumentalism?
https://karpathy.ai/lexicap/0005-large.html#00:02:36.280
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
For predictions.
https://karpathy.ai/lexicap/0005-large.html#00:02:38.440
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
If you believe that this is law of God,
https://karpathy.ai/lexicap/0005-large.html#00:02:39.960
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and it's always true everywhere,
https://karpathy.ai/lexicap/0005-large.html#00:02:43.680
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
that means that you're realist.
https://karpathy.ai/lexicap/0005-large.html#00:02:47.560
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
So you're trying to really understand God's thought.
https://karpathy.ai/lexicap/0005-large.html#00:02:50.080
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
So the way you see the world is as an instrumentalist?
https://karpathy.ai/lexicap/0005-large.html#00:02:55.520
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
You know, I'm working for some models,
https://karpathy.ai/lexicap/0005-large.html#00:03:00.080
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
model of machine learning.
https://karpathy.ai/lexicap/0005-large.html#00:03:03.280
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
So in this model, we can see setting,
https://karpathy.ai/lexicap/0005-large.html#00:03:07.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and we try to solve,
https://karpathy.ai/lexicap/0005-large.html#00:03:12.840
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
resolve the setting to solve the problem.
https://karpathy.ai/lexicap/0005-large.html#00:03:15.360
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
And you can do in two different way.
https://karpathy.ai/lexicap/0005-large.html#00:03:18.320
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
From the point of view of instrumentalist,
https://karpathy.ai/lexicap/0005-large.html#00:03:20.840
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and that's what everybody does now.
https://karpathy.ai/lexicap/0005-large.html#00:03:23.880
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Because they say that goal of machine learning
https://karpathy.ai/lexicap/0005-large.html#00:03:27.160
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
is to find the rule for classification.
https://karpathy.ai/lexicap/0005-large.html#00:03:31.640
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
That is true.
https://karpathy.ai/lexicap/0005-large.html#00:03:36.880
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
But it is instrument for prediction.
https://karpathy.ai/lexicap/0005-large.html#00:03:38.360
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
But I can say the goal of machine learning
https://karpathy.ai/lexicap/0005-large.html#00:03:41.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
is to learn about conditional probability.
https://karpathy.ai/lexicap/0005-large.html#00:03:46.240
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
So how God played use, and if he play,
https://karpathy.ai/lexicap/0005-large.html#00:03:50.080
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
what is probability for one,
https://karpathy.ai/lexicap/0005-large.html#00:03:54.520
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
what is probability for another, given situation.
https://karpathy.ai/lexicap/0005-large.html#00:03:56.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
But for prediction, I don't need this.
https://karpathy.ai/lexicap/0005-large.html#00:04:00.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
I need the rule.
https://karpathy.ai/lexicap/0005-large.html#00:04:02.680
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
But for understanding, I need conditional probability.
https://karpathy.ai/lexicap/0005-large.html#00:04:04.320
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
So let me just step back a little bit first to talk about,
https://karpathy.ai/lexicap/0005-large.html#00:04:08.520
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
you mentioned, which I read last night,
https://karpathy.ai/lexicap/0005-large.html#00:04:11.840
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
the parts of the 1960 paper by Eugene Wigner,
https://karpathy.ai/lexicap/0005-large.html#00:04:14.000
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Unreasonable Effectiveness of Mathematics
https://karpathy.ai/lexicap/0005-large.html#00:04:21.360
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
and Natural Sciences.
https://karpathy.ai/lexicap/0005-large.html#00:04:23.560
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Such a beautiful paper, by the way.
https://karpathy.ai/lexicap/0005-large.html#00:04:24.960
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Made me feel, to be honest,
https://karpathy.ai/lexicap/0005-large.html#00:04:29.400
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
to confess my own work in the past few years
https://karpathy.ai/lexicap/0005-large.html#00:04:32.640
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
on deep learning, heavily applied.
https://karpathy.ai/lexicap/0005-large.html#00:04:35.560
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Made me feel that I was missing out
https://karpathy.ai/lexicap/0005-large.html#00:04:38.480
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
on some of the beauty of nature
https://karpathy.ai/lexicap/0005-large.html#00:04:40.440
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
in the way that math can uncover.
https://karpathy.ai/lexicap/0005-large.html#00:04:43.480
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
So let me just step away from the poetry of that for a second.
https://karpathy.ai/lexicap/0005-large.html#00:04:45.640
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
How do you see the role of math in your life?
https://karpathy.ai/lexicap/0005-large.html#00:04:50.440
Vladimir Vapnik: Statistical Learning | Lex Fridman Podcast #5
Is it a tool, is it poetry?
https://karpathy.ai/lexicap/0005-large.html#00:04:53.120