episode
stringlengths
45
100
text
stringlengths
1
528
timestamp_link
stringlengths
56
56
Guido van Rossum: Python | Lex Fridman Podcast #6
But people who are more interested in emphasizing the object oriented nature of things.
https://karpathy.ai/lexicap/0006-large.html#00:58:21.200
Guido van Rossum: Python | Lex Fridman Podcast #6
So within the last 10 years or so, especially with the oncoming of neural networks and these
https://karpathy.ai/lexicap/0006-large.html#00:58:29.160
Guido van Rossum: Python | Lex Fridman Podcast #6
packages that are built on Python to interface with neural networks, I switched to Python
https://karpathy.ai/lexicap/0006-large.html#00:58:34.920
Guido van Rossum: Python | Lex Fridman Podcast #6
and it's just, I've noticed a significant boost that I can't exactly, because I don't
https://karpathy.ai/lexicap/0006-large.html#00:58:41.360
Guido van Rossum: Python | Lex Fridman Podcast #6
think about it, but I can't exactly put into words why I'm just much, much more productive.
https://karpathy.ai/lexicap/0006-large.html#00:58:47.120
Guido van Rossum: Python | Lex Fridman Podcast #6
Just being able to get the job done much, much faster.
https://karpathy.ai/lexicap/0006-large.html#00:58:52.840
Guido van Rossum: Python | Lex Fridman Podcast #6
So how do you think, whatever that qualitative difference is, I don't know if it's quantitative,
https://karpathy.ai/lexicap/0006-large.html#00:58:56.400
Guido van Rossum: Python | Lex Fridman Podcast #6
it could be just a feeling, I don't know if I'm actually more productive, but how
https://karpathy.ai/lexicap/0006-large.html#00:59:01.880
Guido van Rossum: Python | Lex Fridman Podcast #6
do you think about...
https://karpathy.ai/lexicap/0006-large.html#00:59:07.280
Guido van Rossum: Python | Lex Fridman Podcast #6
You probably are.
https://karpathy.ai/lexicap/0006-large.html#00:59:08.280
Guido van Rossum: Python | Lex Fridman Podcast #6
Yeah.
https://karpathy.ai/lexicap/0006-large.html#00:59:09.280
Guido van Rossum: Python | Lex Fridman Podcast #6
Well, that's right.
https://karpathy.ai/lexicap/0006-large.html#00:59:10.280
Guido van Rossum: Python | Lex Fridman Podcast #6
I think there's elements, let me just speak to one aspect that I think that was affecting
https://karpathy.ai/lexicap/0006-large.html#00:59:11.880
Guido van Rossum: Python | Lex Fridman Podcast #6
my productivity is C++ was, I really enjoyed creating performant code and creating a beautiful
https://karpathy.ai/lexicap/0006-large.html#00:59:15.400
Guido van Rossum: Python | Lex Fridman Podcast #6
structure where everything that, you know, this kind of going into this, especially with
https://karpathy.ai/lexicap/0006-large.html#00:59:26.160
Guido van Rossum: Python | Lex Fridman Podcast #6
the newer and newer standards of templated programming of just really creating this beautiful
https://karpathy.ai/lexicap/0006-large.html#00:59:31.000
Guido van Rossum: Python | Lex Fridman Podcast #6
formal structure that I found myself spending most of my time doing that as opposed to getting
https://karpathy.ai/lexicap/0006-large.html#00:59:37.080
Guido van Rossum: Python | Lex Fridman Podcast #6
it, parsing a file and extracting a few keywords or whatever the task was trying to do.
https://karpathy.ai/lexicap/0006-large.html#00:59:42.000
Guido van Rossum: Python | Lex Fridman Podcast #6
So what is it about Python?
https://karpathy.ai/lexicap/0006-large.html#00:59:47.520
Guido van Rossum: Python | Lex Fridman Podcast #6
How do you think of productivity in general as you were designing it now, sort of through
https://karpathy.ai/lexicap/0006-large.html#00:59:49.980
Guido van Rossum: Python | Lex Fridman Podcast #6
the decades, last three decades, what do you think it means to be a productive programmer?
https://karpathy.ai/lexicap/0006-large.html#00:59:54.520
Guido van Rossum: Python | Lex Fridman Podcast #6
And how did you try to design it into the language?
https://karpathy.ai/lexicap/0006-large.html#01:00:00.120
Guido van Rossum: Python | Lex Fridman Podcast #6
There are different tasks and as a programmer, it's useful to have different tools available
https://karpathy.ai/lexicap/0006-large.html#01:00:03.560
Guido van Rossum: Python | Lex Fridman Podcast #6
that sort of are suitable for different tasks.
https://karpathy.ai/lexicap/0006-large.html#01:00:10.400
Guido van Rossum: Python | Lex Fridman Podcast #6
So I still write C code, I still write shell code, but I write most of my things in Python.
https://karpathy.ai/lexicap/0006-large.html#01:00:13.940
Guido van Rossum: Python | Lex Fridman Podcast #6
Why do I still use those other languages, because sometimes the task just demands it.
https://karpathy.ai/lexicap/0006-large.html#01:00:25.600
Guido van Rossum: Python | Lex Fridman Podcast #6
And well, I would say most of the time the task actually demands a certain language because
https://karpathy.ai/lexicap/0006-large.html#01:00:33.000
Guido van Rossum: Python | Lex Fridman Podcast #6
the task is not write a program that solves problem X from scratch, but it's more like
https://karpathy.ai/lexicap/0006-large.html#01:00:39.000
Guido van Rossum: Python | Lex Fridman Podcast #6
fix a bug in existing program X or add a small feature to an existing large program.
https://karpathy.ai/lexicap/0006-large.html#01:00:45.600
Guido van Rossum: Python | Lex Fridman Podcast #6
But even if you're not constrained in your choice of language by context like that, there
https://karpathy.ai/lexicap/0006-large.html#01:00:56.680
Guido van Rossum: Python | Lex Fridman Podcast #6
is still the fact that if you write it in a certain language, then you have this balance
https://karpathy.ai/lexicap/0006-large.html#01:01:10.160
Guido van Rossum: Python | Lex Fridman Podcast #6
between how long does it take you to write the code and how long does the code run?
https://karpathy.ai/lexicap/0006-large.html#01:01:21.360
Guido van Rossum: Python | Lex Fridman Podcast #6
And when you're in the phase of exploring solutions, you often spend much more time
https://karpathy.ai/lexicap/0006-large.html#01:01:31.840
Guido van Rossum: Python | Lex Fridman Podcast #6
writing the code than running it because every time you've run it, you see that the output
https://karpathy.ai/lexicap/0006-large.html#01:01:42.760
Guido van Rossum: Python | Lex Fridman Podcast #6
is not quite what you wanted and you spend some more time coding.
https://karpathy.ai/lexicap/0006-large.html#01:01:50.720
Guido van Rossum: Python | Lex Fridman Podcast #6
And a language like Python just makes that iteration much faster because there are fewer
https://karpathy.ai/lexicap/0006-large.html#01:01:58.480
Guido van Rossum: Python | Lex Fridman Podcast #6
details that you have to get right before your program compiles and runs.
https://karpathy.ai/lexicap/0006-large.html#01:02:06.760
Guido van Rossum: Python | Lex Fridman Podcast #6
There are libraries that do all sorts of stuff for you, so you can sort of very quickly take
https://karpathy.ai/lexicap/0006-large.html#01:02:19.480
Guido van Rossum: Python | Lex Fridman Podcast #6
a bunch of existing components, put them together, and get your prototype application running.
https://karpathy.ai/lexicap/0006-large.html#01:02:26.400
Guido van Rossum: Python | Lex Fridman Podcast #6
Just like when I was building electronics, I was using a breadboard most of the time,
https://karpathy.ai/lexicap/0006-large.html#01:02:36.320
Guido van Rossum: Python | Lex Fridman Podcast #6
so I had this sprawl out circuit that if you shook it, it would stop working because it
https://karpathy.ai/lexicap/0006-large.html#01:02:42.860
Guido van Rossum: Python | Lex Fridman Podcast #6
was not put together very well, but it functioned and all I wanted was to see that it worked
https://karpathy.ai/lexicap/0006-large.html#01:02:51.320
Guido van Rossum: Python | Lex Fridman Podcast #6
and then move on to the next schematic or design or add something to it.
https://karpathy.ai/lexicap/0006-large.html#01:02:58.800
Guido van Rossum: Python | Lex Fridman Podcast #6
Once you've sort of figured out, oh, this is the perfect design for my radio or light
https://karpathy.ai/lexicap/0006-large.html#01:03:05.000
Guido van Rossum: Python | Lex Fridman Podcast #6
sensor or whatever, then you can say, okay, how do we design a PCB for this?
https://karpathy.ai/lexicap/0006-large.html#01:03:10.500
Guido van Rossum: Python | Lex Fridman Podcast #6
How do we solder the components in a small space?
https://karpathy.ai/lexicap/0006-large.html#01:03:15.800
Guido van Rossum: Python | Lex Fridman Podcast #6
How do we make it so that it is robust against, say, voltage fluctuations or mechanical disruption?
https://karpathy.ai/lexicap/0006-large.html#01:03:19.920
Guido van Rossum: Python | Lex Fridman Podcast #6
I know nothing about that when it comes to designing electronics, but I know a lot about
https://karpathy.ai/lexicap/0006-large.html#01:03:32.840
Guido van Rossum: Python | Lex Fridman Podcast #6
that when it comes to writing code.
https://karpathy.ai/lexicap/0006-large.html#01:03:37.320
Guido van Rossum: Python | Lex Fridman Podcast #6
So the initial steps are efficient, fast, and there's not much stuff that gets in the
https://karpathy.ai/lexicap/0006-large.html#01:03:40.400
Guido van Rossum: Python | Lex Fridman Podcast #6
way, but you're kind of describing, like Darwin described the evolution of species, right?
https://karpathy.ai/lexicap/0006-large.html#01:03:46.080
Guido van Rossum: Python | Lex Fridman Podcast #6
You're observing of what is true about Python.
https://karpathy.ai/lexicap/0006-large.html#01:03:56.680
Guido van Rossum: Python | Lex Fridman Podcast #6
Now if you take a step back, if the act of creating languages is art and you had three
https://karpathy.ai/lexicap/0006-large.html#01:04:00.520
Guido van Rossum: Python | Lex Fridman Podcast #6
months to do it, initial steps, so you just specified a bunch of goals, sort of things
https://karpathy.ai/lexicap/0006-large.html#01:04:07.800
Guido van Rossum: Python | Lex Fridman Podcast #6
that you observe about Python, perhaps you had those goals, but how do you create the
https://karpathy.ai/lexicap/0006-large.html#01:04:15.640
Guido van Rossum: Python | Lex Fridman Podcast #6
rules, the syntactic structure, the features that result in those?
https://karpathy.ai/lexicap/0006-large.html#01:04:19.400
Guido van Rossum: Python | Lex Fridman Podcast #6
So I have in the beginning and I have follow up questions about through the evolution of
https://karpathy.ai/lexicap/0006-large.html#01:04:25.600
Guido van Rossum: Python | Lex Fridman Podcast #6
Python too, but in the very beginning when you were sitting there creating the lexical
https://karpathy.ai/lexicap/0006-large.html#01:04:29.880
Guido van Rossum: Python | Lex Fridman Podcast #6
analyzer or whatever.
https://karpathy.ai/lexicap/0006-large.html#01:04:35.440
Guido van Rossum: Python | Lex Fridman Podcast #6
Python was still a big part of it because I sort of, I said to myself, I don't want
https://karpathy.ai/lexicap/0006-large.html#01:04:37.440
Guido van Rossum: Python | Lex Fridman Podcast #6
to have to design everything from scratch, I'm going to borrow features from other languages
https://karpathy.ai/lexicap/0006-large.html#01:04:47.240
Guido van Rossum: Python | Lex Fridman Podcast #6
that I like.
https://karpathy.ai/lexicap/0006-large.html#01:04:53.640
Guido van Rossum: Python | Lex Fridman Podcast #6
Oh, interesting.
https://karpathy.ai/lexicap/0006-large.html#01:04:54.640
Guido van Rossum: Python | Lex Fridman Podcast #6
So you basically, exactly, you first observe what you like.
https://karpathy.ai/lexicap/0006-large.html#01:04:55.640
Guido van Rossum: Python | Lex Fridman Podcast #6
Yeah, and so that's why if you're 17 years old and you want to sort of create a programming
https://karpathy.ai/lexicap/0006-large.html#01:04:58.360
Guido van Rossum: Python | Lex Fridman Podcast #6
language, you're not going to be very successful at it because you have no experience with
https://karpathy.ai/lexicap/0006-large.html#01:05:05.240
Guido van Rossum: Python | Lex Fridman Podcast #6
other languages, whereas I was in my, let's say mid 30s, I had written parsers before,
https://karpathy.ai/lexicap/0006-large.html#01:05:11.600
Guido van Rossum: Python | Lex Fridman Podcast #6
so I had worked on the implementation of ABC, I had spent years debating the design of ABC
https://karpathy.ai/lexicap/0006-large.html#01:05:24.300
Guido van Rossum: Python | Lex Fridman Podcast #6
with its authors, with its designers, I had nothing to do with the design, it was designed
https://karpathy.ai/lexicap/0006-large.html#01:05:30.880
Guido van Rossum: Python | Lex Fridman Podcast #6
fully as it ended up being implemented when I joined the team.
https://karpathy.ai/lexicap/0006-large.html#01:05:37.520
Guido van Rossum: Python | Lex Fridman Podcast #6
But so you borrow ideas and concepts and very concrete sort of local rules from different
https://karpathy.ai/lexicap/0006-large.html#01:05:42.080
Guido van Rossum: Python | Lex Fridman Podcast #6
languages like the indentation and certain other syntactic features from ABC, but I chose
https://karpathy.ai/lexicap/0006-large.html#01:05:51.440
Guido van Rossum: Python | Lex Fridman Podcast #6
to borrow string literals and how numbers work from C and various other things.
https://karpathy.ai/lexicap/0006-large.html#01:05:58.920
Guido van Rossum: Python | Lex Fridman Podcast #6
So in then, if you take that further, so yet you've had this funny sounding, but I think
https://karpathy.ai/lexicap/0006-large.html#01:06:07.960
Guido van Rossum: Python | Lex Fridman Podcast #6
surprisingly accurate and at least practical title of benevolent dictator for life for
https://karpathy.ai/lexicap/0006-large.html#01:06:13.800
Guido van Rossum: Python | Lex Fridman Podcast #6
quite, you know, for the last three decades or whatever, or no, not the actual title,
https://karpathy.ai/lexicap/0006-large.html#01:06:21.000
Guido van Rossum: Python | Lex Fridman Podcast #6
but functionally speaking.
https://karpathy.ai/lexicap/0006-large.html#01:06:25.240
Guido van Rossum: Python | Lex Fridman Podcast #6
So you had to make decisions, design decisions.
https://karpathy.ai/lexicap/0006-large.html#01:06:27.940
Guido van Rossum: Python | Lex Fridman Podcast #6
Can you maybe, let's take Python 2, so releasing Python 3 as an example.
https://karpathy.ai/lexicap/0006-large.html#01:06:34.280
Guido van Rossum: Python | Lex Fridman Podcast #6
It's not backward compatible to Python 2 in ways that a lot of people know.
https://karpathy.ai/lexicap/0006-large.html#01:06:41.960
Guido van Rossum: Python | Lex Fridman Podcast #6
So what was that deliberation, discussion, decision like?
https://karpathy.ai/lexicap/0006-large.html#01:06:47.240
Guido van Rossum: Python | Lex Fridman Podcast #6
Yeah.
https://karpathy.ai/lexicap/0006-large.html#01:06:50.640
Guido van Rossum: Python | Lex Fridman Podcast #6
What was the psychology of that experience?
https://karpathy.ai/lexicap/0006-large.html#01:06:51.640
Guido van Rossum: Python | Lex Fridman Podcast #6
Do you regret any aspects of how that experience undergone that?
https://karpathy.ai/lexicap/0006-large.html#01:06:54.520
Guido van Rossum: Python | Lex Fridman Podcast #6
Well, yeah, so it was a group process really.
https://karpathy.ai/lexicap/0006-large.html#01:06:58.520
Guido van Rossum: Python | Lex Fridman Podcast #6
At that point, even though I was BDFL in name and certainly everybody sort of respected
https://karpathy.ai/lexicap/0006-large.html#01:07:03.040
Guido van Rossum: Python | Lex Fridman Podcast #6
my position as the creator and the current sort of owner of the language design, I was
https://karpathy.ai/lexicap/0006-large.html#01:07:11.880
Guido van Rossum: Python | Lex Fridman Podcast #6
looking at everyone else for feedback.
https://karpathy.ai/lexicap/0006-large.html#01:07:22.160
Guido van Rossum: Python | Lex Fridman Podcast #6
Sort of Python 3.0 in some sense was sparked by other people in the community pointing
https://karpathy.ai/lexicap/0006-large.html#01:07:26.560
Guido van Rossum: Python | Lex Fridman Podcast #6
out, oh, well, there are a few issues that sort of bite users over and over.
https://karpathy.ai/lexicap/0006-large.html#01:07:35.280
Guido van Rossum: Python | Lex Fridman Podcast #6
Can we do something about that?
https://karpathy.ai/lexicap/0006-large.html#01:07:46.360
Guido van Rossum: Python | Lex Fridman Podcast #6
And for Python 3, we took a number of those Python words as they were called at the time
https://karpathy.ai/lexicap/0006-large.html#01:07:48.920
Guido van Rossum: Python | Lex Fridman Podcast #6
and we said, can we try to sort of make small changes to the language that address those
https://karpathy.ai/lexicap/0006-large.html#01:07:56.360
Guido van Rossum: Python | Lex Fridman Podcast #6
words?
https://karpathy.ai/lexicap/0006-large.html#01:08:04.800
Guido van Rossum: Python | Lex Fridman Podcast #6
And we had sort of in the past, we had always taken backwards compatibility very seriously.
https://karpathy.ai/lexicap/0006-large.html#01:08:06.560
Guido van Rossum: Python | Lex Fridman Podcast #6
And so many Python words in earlier versions had already been resolved because they could
https://karpathy.ai/lexicap/0006-large.html#01:08:15.360
Guido van Rossum: Python | Lex Fridman Podcast #6
be resolved while maintaining backwards compatibility or sort of using a very gradual path of evolution
https://karpathy.ai/lexicap/0006-large.html#01:08:20.420
Guido van Rossum: Python | Lex Fridman Podcast #6
of the language in a certain area.
https://karpathy.ai/lexicap/0006-large.html#01:08:29.740
Guido van Rossum: Python | Lex Fridman Podcast #6
And so we were stuck with a number of words that were widely recognized as problems, not
https://karpathy.ai/lexicap/0006-large.html#01:08:31.960
Guido van Rossum: Python | Lex Fridman Podcast #6
like roadblocks, but nevertheless sort of things that some people trip over and you know that
https://karpathy.ai/lexicap/0006-large.html#01:08:39.760